Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The formation of 6-azaspirocycles via semipinacol rearrangement reactions and their application in a… Hurley, Paul 2006

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

Item Metadata

Download

Media
831-ubc_2006_fall_hurley_paul.pdf [ 16.49MB ]
Metadata
JSON: 831-1.0061118.json
JSON-LD: 831-1.0061118-ld.json
RDF/XML (Pretty): 831-1.0061118-rdf.xml
RDF/JSON: 831-1.0061118-rdf.json
Turtle: 831-1.0061118-turtle.txt
N-Triples: 831-1.0061118-rdf-ntriples.txt
Original Record: 831-1.0061118-source.json
Full Text
831-1.0061118-fulltext.txt
Citation
831-1.0061118.ris

Full Text

THE FORMATION OF 6-AZASPIROCYCLES VIA SEMIPINACOL REARRANGEMENT REACTIONS AND THEIR APPLICATION IN A SYNTHETIC ROUTE TOWARDS HALICHLORINE  by PAUL HURLEY B. Sc. (Hons.), Memorial University of Newfoundland, 2000  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA August 2006 © Paul Hurley, 2006  ii Abstract This document describes a synthetic approach towards the tricyclic unit contained within the natural product halichlorine based upon a semipinacol rearrangement reaction as a key transformation. A number of synthetic approaches involving the synthesis of halichlorine and the structurally related compounds pinnaic acid and tauropinnaic acid have been published; this work is described in chapter 1. In chapter 2, a detailed account of our first approach towards the tricyclic core of halichlorine is described. This approach involves formation of one of the rings of halichlorine by a ring closing metathesis reaction. To achieve this goal, a new, modified version of Grubbs’ “second generation” ring closing metathesis catalyst was synthesized. This catalyst exhibits high reactivity and successfully closed a 6-membered ring in a compound that contains structural features similar to those found in halichlorine. Our approach towards the synthesis of the tricyclic core of halichlorine led to the development of a new method to form 6-azaspirocyclopentanones. When piperidine-based allylic cyclobutanols are treated with N-bromosuccinimide, a ring expansion reaction takes place that results in the formation of highly functionalized 6-azaspirocyclopentanones. These high yielding, diastereoselective reactions were successful with several ring expansion substrates. The synthesis of the ring expansion substrates led to the development of a new method to construct alkenyl stannanes from isolated enol triflates using lithium trimethylstannyl copper (I) cyanide reagent. The semipinacol rearrangement reactions outlined in chapter 2 gave products with the incorrect relative configuration required for halichlorine. These results led to the development and implementation of a new asymmetric synthetic sequence towards the tricyclic core of halichlorine that is discussed in chapter 3. This synthetic sequence involves the N-bromosuccinimide promoted ring expansion reaction of a piperidine-based allylic cyclobutanol that contains a substituent on the cyclobutane ring. This ring expansion reaction resulted in the formation of a densely functionalized azaspirocyclopentanone that contains four of the five stereocenters and two of the four rings required to make halichlorine. Ultimately a late stage intermediate was achieved in 22 steps (longest linear sequence) from 1,3-propanediol.  iii Table of Contents Abstract............................................................................................................................................. ii Table of Contents............................................................................................................................. iii List of Tables ................................................................................................................................... xi List of Figures................................................................................................................................ xiv List of Schemes.............................................................................................................................. xvi List of Symbols and Abbreviations ................................................................................................ xx Acknowledgements....................................................................................................................... xxv Foreword...................................................................................................................................... xxvi 1  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids ........................ 1 1.1  Introduction ...................................................................................................................... 2  1.2  Biological Activity............................................................................................................ 2  1.3  Synthetic Analysis ............................................................................................................ 3  1.4  Synthetic Approaches Towards Halichlorine and the Pinnaic Acids ............................... 4  1.4.1  Introduction .................................................................................................................. 4  1.4.2  Groups that Start with the C Ring ................................................................................ 5  1.4.2.1  Contributions from the Danishefsky Laboratory: Total Synthesis of  Halichlorine…… ..................................................................................................................... 5 1.4.2.2  Contributions from the Danishefsky Laboratory: Synthesis of Pinnaic Acid ...... 6  1.4.2.3  The Kibayashi Approach Towards Pinnaic Acid ................................................. 8  1.4.2.4  Kibayashi’s Completion of Halichlorine .............................................................. 9  1.4.2.5  Heathcock’s Synthesis of Pinnaic Acid and Tauropinnaic Acid .......................... 9  1.4.2.6  Heathcock’s Completion of Halichlorine ........................................................... 11  1.4.2.7  Uemura/Arimoto Approach Towards Pinnaic Acid ........................................... 12  1.4.2.8  Uemura’s Synthesis of the Tricyclic Core of Halichlorine ................................ 13  1.4.2.9  Contributions from the Zhao and Ding Laboratories ......................................... 14  1.4.2.10  Contributions from the Martin Laboratory ......................................................... 15  1.4.2.11  The Martin Approach Towards Halichlorine ..................................................... 16  1.4.2.12  Contributions from Pilli’s Laboratory ................................................................ 16  1.4.2.13  Contributions from the Forsyth Laboratory........................................................ 17  1.4.2.14  Contributions from the Wright Laboratory......................................................... 18  1.4.3  Groups That Start with the B Ring ............................................................................. 19  1.4.3.1  iv Clive’s First Generation Approach Towards Halichlorine................................. 19  1.4.3.2  Clive’s Second Generation Approach Towards Halichlorine and the Pinnaic  Acids  ............................................................................................................................ 20  1.4.3.3  Clive’s Third Generation Approach Towards Halichlorine ............................... 21  1.4.3.4  The Simpkins First Generation Approach Towards Halichlorine ...................... 22  1.4.3.5  The Simpkins Second Generation Approach Towards Halichlorine.................. 22  1.4.3.6  Contributions from the White Group.................................................................. 23  1.4.3.7  Contributions from the Feldman Group ............................................................. 24  1.4.3.8  Contributions from the Ihara Laboratory............................................................ 25  1.4.3.9  Contributions from the Laboratories of Mol and Bubnov .................................. 26  1.4.3.10  Contributions from the Keck Laboratory ........................................................... 27  1.4.4  Groups that Make the B and C Rings in One Step ..................................................... 27  1.4.4.1  The Shishido/Itoh First Generation Approach Towards Halichlorine................ 27  1.4.4.2  The Shishido/Itoh Second Generation Approach Towards Halichlorine ........... 28  1.4.4.3  Contributions from the Zhao Laboratories ......................................................... 29  1.4.4.4  Contributions from the Stockman Laboratory.................................................... 30  1.4.5  Groups That Synthesize the C15-C21 Side Chain ........................................................ 30  1.4.5.1  Contributions from the Weinreb Laboratory ...................................................... 30  1.4.5.2  Studies from the Taber Laboratories .................................................................. 31  1.5  Conclusion ...................................................................................................................... 31  1.6  References....................................................................................................................... 33  2  Chapter 2 A First Generation Approach Towards Halichlorine and the Development  of New Methods to form 6-Azaspirocycles ................................................................................. 35 2.1  Introduction .................................................................................................................... 36  2.2  Retrosynthetic Analysis .................................................................................................. 36  2.3  Previous Work in the Dake Lab...................................................................................... 38  2.4  Formation of the A Ring of Halichlorine by Ring Closing Metathesis: Substrate  Synthesis ...................................................................................................................................... 43 2.4.1  Introduction ................................................................................................................ 43  2.4.2  Path A: Synthesis of Dienes 2.1.6a and 2.1.6b .......................................................... 44  2.5 2.5.1  Metathesis ....................................................................................................................... 45 Introduction ................................................................................................................ 45  2.5.2  v Types of Metathesis Reactions ................................................................................... 46  2.5.3  Important Advances in Metathesis Catalyst Design................................................... 47  2.5.4  Mechanistic Considerations........................................................................................ 50  2.5.5  Summary..................................................................................................................... 51  2.6  Formation of the A Ring of Halichlorine by Ring Closing Metathesis: Ring Closing  Metathesis and Elaboration of Diene 2.1.6a................................................................................ 52 2.7  Formation of the A Ring of Halichlorine by Ring Closing Metathesis: Selection and  Synthesis of a Second Generation Metathesis Catalyst............................................................... 54 2.8  Ring Closing Metathesis and Attempted Elaboration of Diene 2.1.5b .......................... 57  2.9  Progress Towards the Tricyclic Core of Halichlorine Via Path B.................................. 59  2.9.1  Introduction ................................................................................................................ 59  2.9.2  Synthesis of Alkenyl Stannane 2.1.11 ........................................................................ 60  2.9.3  Synthesis of the Semipinacol Substrate 2.1.4............................................................. 60  2.9.4  Attempted Acid Catalyzed Semipinacol Rearrangement of 2.1.10............................ 61  2.9.5  Can the Ethoxy Group be Replaced by an Allyl Group?............................................ 62  2.9.6  Summary..................................................................................................................... 64  2.10  Electrophilic Halogenation Reactions of Alkenes.......................................................... 65  2.10.1  Introduction ............................................................................................................ 65  2.10.2  Reactions of Alkenes with Diatomic Halogen Reagents........................................ 65  2.10.3  Mechanism.............................................................................................................. 66  2.10.4  Reactions of Simple Alkenes where the Electrophilic Halogen Source and the  Nucleophile are Different ......................................................................................................... 67 2.10.5  Diastereoselective Reactions of Chiral Cyclic Alkenes ......................................... 68  2.10.6  Diastereoselective Reactions of Cycloalkenes that Contain Alcohols/Ethers........ 69  2.10.7  Reactions of Glycals ............................................................................................... 71  2.10.8  Electrophilic Halogen Promoted Ring Expansion Reactions of Cycloalkyl Allylic  Alcohols  ................................................................................................................................ 75  2.10.9  Summary................................................................................................................. 79  2.11  An Electrophilic Bromine Solution to the Semipinacol Rearrangement Reaction......... 80  2.11.1  Attempted Semipinacol Rearrangement Reactions with Allylic Alcohol 2.1.10 ... 80  2.11.2  Determination of the Stereochemical Configuration of 2.29.1............................... 81  2.11.3  Mechanistic Rationale to Explain the Formation of Spirocyclopentanone 2.29.1 . 81  2.12  Synthesis of Substrates to be used in the N-bromosuccinimide Promoted Ring  vi Expansion Reaction ..................................................................................................................... 82 2.12.1  Introduction ............................................................................................................ 82  2.12.2  Potential Substrates for the N-bromosuccinimide Promoted Ring Expansion  Reaction  ................................................................................................................................ 83  2.12.3  Retrosynthetic Analysis of Semipinacol Rearrangement Substrates...................... 84  2.12.4  Amide Preparation .................................................................................................. 85  2.12.5  Alkenyl Triflate Formation..................................................................................... 88  2.12.6  Alkenyl Stannane Formation .................................................................................. 88  2.12.7  Alternative Alkenyl Stannane Procedure................................................................ 89  2.12.8  Synthesis of the Allylic Cyclobutanols................................................................... 91  2.12.9  Summary................................................................................................................. 92  2.13  N-Bromosuccinimide Promoted Ring Expansion Reactions.......................................... 93  2.13.1  Stereochemical Assignments .................................................................................. 95  2.13.1.1  Introduction ........................................................................................................ 95  2.13.1.2  Substituent at the 4-Position ............................................................................... 95  2.13.1.3  Substituent at the 5-position ............................................................................... 97  2.13.1.4  Substituent at the 6-position ............................................................................. 100  2.13.2  Rationalization of Diastereoselectivity................................................................. 100  2.13.2.1  Substituent at the 4-Position ............................................................................. 100  2.13.2.2  Substituent at the 5-Position ............................................................................. 101  2.13.2.3  Substitution at the 6-Position............................................................................ 102  2.13.3  Summary............................................................................................................... 103  2.14  Additional Substrates for the Acid Catalyzed Semipinacol Rearrangement and the  Verification of the Configuration of the Products ..................................................................... 103 2.14.1  Introduction .......................................................................................................... 103  2.14.2  Acid Catalyzed Sempinacol Rearrangements....................................................... 104  2.14.3  Stereochemical Assignments ................................................................................ 105  2.14.3.1  Substituent at the 4-Position ............................................................................. 105  2.14.3.2  The 6-Allyl Product 2.41.2a ............................................................................. 108  2.14.4 2.15  Summary............................................................................................................... 109 The Halichlorine Synthesis Revisited........................................................................... 109  2.15.1  Introduction .......................................................................................................... 109  2.15.2  Synthetic Elaboration of 2.29.1 ............................................................................ 110  vii Siloxy-epoxide Approach Towards Halichlorine ................................................. 111  2.15.3 2.16  Chapter Summary ......................................................................................................... 114  2.17  Experimental................................................................................................................. 116  2.18  References..................................................................................................................... 191  3  Chapter 3 An Asymmetric Substituted Cyclobutanol-based Approach Towards the  Synthesis of Halichlorine............................................................................................................ 201 3.1  Introduction .................................................................................................................. 202  3.2  Revised Retrosynthetic Analysis .................................................................................. 202  3.3  Points of Concern ......................................................................................................... 203  3.4  Synthesis of Substituted Cyclobutanones..................................................................... 206  3.4.1  Alkylation of Hydrazone .......................................................................................... 206  3.4.2  Organometallic-based Approaches Towards Substituted Cyclobutanones .............. 206  3.4.2.1  Bergman’s Synthesis of 2-Substituted Cyclobutanones from  Cobaltocyclopentanones ...................................................................................................... 206 3.4.2.2  Substituted Cyclobutanones from Titanacyclobutanes..................................... 207  3.4.3  TosMic Approach Towards 2-Substituted Cyclobutanones ..................................... 208  3.4.4  [2+2] Cycloaddition Approach Towards Substituted Cyclobutanones .................... 209  3.4.5  Synthesis of 2-Alkoxycyclobutanones ..................................................................... 209  3.4.6  Substituted Cyclobutanones via Ring Expansion Reactions of Cyclopropanes....... 210  3.4.6.1  Synthesis of 2-Alkenyl Cyclobutanones........................................................... 210  3.4.6.2  Synthesis of Substituted Cyclobutanones From Cyclopropylphenylsulfide .... 211  3.4.6.3  Gadwood’s Synthesis of Substituted Cyclobutanones ..................................... 211  3.4.6.4  Cyclobutanones via Epoxidation or Dihydroxylation of Cyclopropylidenes... 211  3.4.6.5  Kulinkovich Cyclopropanation Approach Towards Substituted Cyclobutanones .......................................................................................................................... 212  3.4.7  Chiral Non-racemic Cyclobutanones........................................................................ 213  3.4.7.1  Bergman’s Synthesis of Non-racemic Cyclobutanones ................................... 213  3.4.7.2  Salaün’s Synthesis of Non-racemic Cyclobutanones ....................................... 213  3.4.7.3  Optically Active Cyclobutanones via Asymmetric Epoxidation Reactions ..... 214  3.4.7.4  Asymmetric Dihydroxylation Approach Towards Non-racemic Cyclobutanones.  3.4.7.5  Asymmetric Dihydroxylation-Kulinkovich Cylopropanation Approach  Towards Optically Active Cyclobutanones ......................................................................... 216  viii Summary................................................................................................................... 216  3.4.8 3.5  Synthesis of Chiral Cyclobutanone 3.1.3 Required for the Halichlorine Synthesis..... 216  3.5.1  Synthesis of Chiral Epoxide 3.18.5 .......................................................................... 216  3.5.2  Diastereoselective Opening of Epoxide 3.18.5......................................................... 217  3.5.3  Attempted Oxidation of 3.18.6 ................................................................................. 218  3.5.4  Selective Protection of the Secondary Alcohol ........................................................ 218  3.5.5  Oxidation of Alcohol 3.18.9 ..................................................................................... 219  3.5.6  Completion of the Synthesis of Chiral Cyclobutanone 3.19.4 ................................. 220  3.5.7  Mechanism of Cyclobutanone Formation ................................................................ 221  3.5.8  Summary................................................................................................................... 222  3.6  Carbonyl Addition Reactions Involving Alkenyl Stannane 2.34.19 and Cyclobutanone  3.19.4 ...................................................................................................................................... 222 3.6.1  Introduction .............................................................................................................. 222  3.6.2  Attempted Carbonyl Addition Reaction Under Standard Conditions ...................... 222  3.6.3  Effect of Using Different Alkyllithium Reagents on the Carbonyl Addition Reaction .................................................................................................................................. 223  3.6.4  Solvent Effects on the Carbonyl Addition Reaction................................................. 224  3.6.5  Effect of Using Different Metals in the Carbonyl Addition Reaction...................... 226  3.6.6  The Effect of Temperature on the CeCl3 Promoted Carbonyl Addition Reaction ... 227  3.6.7  The Effect of HMPA on CeCl3 Promoted Carbonyl Addition Reactions ................ 227  3.6.8  The Effect of Using Lewis Acids to Activate Cyclobutanone 3.19.4 in CeCl3  Promoted Carbonyl Addition Reactions................................................................................. 228 3.6.9  The Effect of Adding HMPA in the CeCl3 Promoted Carbonyl Addition Reaction  Where Cyclobutanone 3.19.4 is Activated with a Lewis Acid............................................... 229 3.6.10 3.7  Summary of Carbonyl Addition Reactions........................................................... 230  Carbonyl Addition Reaction with a Model Cyclobutanone ......................................... 230  3.7.1  Introduction .............................................................................................................. 230  3.7.2  Synthesis of Cyclobutanone 3.20.1 .......................................................................... 231  3.7.3  Carbonyl Addition Reactions involving Alkenyl Stannane 2.34.19 and  Cyclobutanone 3.20.1 ............................................................................................................. 232 3.8  Generation of the Organometallic Reagent for the Carbonyl Addition Reaction from an  Alkenyl Iodide ........................................................................................................................... 233 3.8.1  Introduction .............................................................................................................. 233  3.8.2  ix Synthesis of Alkenyl Iodide 3.22.1b ........................................................................ 234  3.8.3  Carbonyl Addition Reactions with Cyclobutanone 3.20.1 ....................................... 235  3.8.4  Stereochemical Assignments .................................................................................... 236  3.8.5  Carbonyl Addition to Chiral Cyclobutanone 3.19.4................................................. 237  3.9  N-Bromosuccinimide Promoted Ring Expansion Reactions of Substituted  Cyclobutanols ............................................................................................................................ 238 3.9.1  Results ...................................................................................................................... 238  3.9.2  Stereochemical Determination.................................................................................. 239  3.9.2.1  Introduction ...................................................................................................... 239  3.9.2.2  Assignments for Spirocyclopentanone 3.23.1 .................................................. 240  3.9.2.3  Assignments for Spirocyclopentanones 3.23.2 and 3.23.3............................... 241  3.10  Elaboration of Spirocyclopentanone 3.23.1.................................................................. 243  3.10.1  Specific Goals that have to be Achieved to Synthesize the Tricyclic Core of  Halichlorine ............................................................................................................................ 243 3.10.2  Attempted Deprotection of the Tosyl Group from Spirocyclopentanone 3.23.1 . 244  3.10.3  Attempted Radical Debromination of Spirocyclopentanone 3.23.1..................... 244  3.10.4  Stereochemical Assignment.................................................................................. 245  3.10.5  Mechanism for the Formation of Vinylogous Lactam 3.24.2 .............................. 246  3.10.6  Reduction of Ketone 3.23.1 .................................................................................. 246  3.10.7  Attempted Deprotection of the Tosyl Group From Alcohol 3.30.4 ..................... 248  3.10.8  Attempted Radical Debromination of Alcohol 3.30.4.......................................... 249  3.10.9  Ozonolysis and Acetal Formation of Terminal Alkene 3.25.4 Prior to Radical  Cleavage of the Bromine ........................................................................................................ 250 3.10.10  Removal of Functional Groups with a Model Compound.................................... 252  3.10.10.1  Introduction .................................................................................................. 252  3.10.10.2  Attempted Wolff-Kischner Reduction.......................................................... 252  3.10.10.3  Attempted Cleavage of Bromide 2.36.8a With Zinc Metal.......................... 253  3.10.10.4  Reduction of Ketone 2.36.8a ........................................................................ 253  3.10.10.5  Attempted Radical Cleavage of Bromide 3.27.4 with Samarium Iodide ..... 253  3.10.10.6  Attempted Debromination of Alcohol 3.27.4 ............................................... 254  3.10.10.7  Attempted E2 Elimination Of Bromide 3.27.4.............................................. 258  3.10.10.8  E2 Elimination Prior to Reduction of Ketone 3.27.4 .................................... 258  3.10.10.9  Attempted Elaboration of Alcohol 3.30.2..................................................... 259  3.10.10.10  x Summary....................................................................................................... 260  3.10.11  E2 Elimination and Reduction of Spirocyclopentanone 3.23.1 ............................ 260  3.10.12  Silylation of Alcohol 3.30.2.................................................................................. 261  3.10.13  Removal of the Tosyl Protecting Group from Bis Silyl ether 3.33.2 ................... 262  3.10.14  Attempted Allylation of Amine 3.33.4 ................................................................. 262  3.10.15  Deprotection and Allylation of Bis Silyl Ether 3.33.4 ......................................... 263  3.10.16  Attempted Ring Closing Metathesis of Acrylate 3.34.1....................................... 264  3.11  Synthesis Summary ...................................................................................................... 265  3.12  Conclusions .................................................................................................................. 266  3.13  Experimental................................................................................................................. 268  3.14  References..................................................................................................................... 315  Afterword...................................................................................................................................... 319 Appendix A Selected Spectra for Chapter 2................................................................................. 321 Appendix B Selected Spectra for Chapter 3 ................................................................................. 365 Appendix C X-Ray Crystallography Data.................................................................................... 399  xi List of Tables Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids None Chapter 2 A First Generation Approach Towards Halichlorine and the Development of New Methods to form 6-Azaspirocycles Table 2. 1 Acid Catalyzed Semipinacol Rearrangement of Allylic Cyclobutanols ..................... 41 Table 2. 2 Attempted Allylation of 2.1.10.................................................................................... 62 Table 2. 3 Attempted N-Tosylation of Lactam 2.6.2.................................................................... 64 Table 2. 4 Reactions of 6-Substituted 2,3-Dihydropyrans ........................................................... 74 Table 2. 5 Formation of Alkenyl Triflates.................................................................................... 88 Table 2. 6 Synthesis of Alkenyl Stannanes .................................................................................. 89 Table 2. 7 Formation of Alkenyl Stannanes Using A Cuprate Reagent....................................... 91 Table 2. 8 Formation of Allylic Cyclobutanols ............................................................................ 92 Table 2. 9 Ring Expansions Promoted by N-Bromosuccinimidea................................................ 94 Table 2. 10 Spectroscopic Data Used to Establish the Structures of 2.36.2a, 2.36.3a and 2.36.4a .............................................................................................................................................. 96 Table 2. 11 Spectroscopic Data Used to Establish the Structures of 2.36.2b, 2.36.3b and 2.36.4b .............................................................................................................................................. 96 Table 2. 12 Conversion of 2.37.2 to 2.36.5a ................................................................................ 99 Table 2. 13 Acid Catalyzed Semipinacol Rearrangement of 2.30.2 and 2.30.6......................... 105 Table 2. 14 Complimentary Diastereoselectivity in the Siloxy-epoxide Semipinacol Rearrangement.................................................................................................................... 112 Table 2. 15 NMR Data for (-)-(5S,8S,10S)-10-bromo-8-(tert-butyldimethylsilyloxy)-6-(toluene4-sulfonyl)-6-azaspiro[4.5]decan-1-one (2.36.6a) ............................................................. 165 Table 2. 16 NMR Data for (-)-(5S,8S,10S)-10-bromo-8-(tert-butyldimethylsilyloxy)-6-(toluene4-sulfonyl)-6-azaspiro[4.5]decan-1-one (2.36.6a) ............................................................. 166 Table 2. 17 NMR Data for (-)-(5R,8S,10R)-10-bromo-8-(tert-butyldimethylsilyloxy)-6-(toluene4-sulfonyl)-6-azaspiro[4.5]decan-1-one (2.36.6b) ............................................................. 167  xii Table 2. 18 COSY Data for (-)-(5R,8S,10R)-10-bromo-8-(tert-butyldimethylsilyloxy)-6-(toluene4-sulfonyl)-6-azaspiro[4.5]decan-1-one (2.36.6b) ............................................................. 168 Table 2. 19 1H Selective NOE Data for (-)-(5R,8S,10R)-10-bromo-8-(tertbutyldimethylsilyloxy)-6-(toluene-4-sulfonyl)-6-azaspiro[4.5]decan-1-one (2.36.6b) ..... 169 Table 2. 20 NMR Data for (+)-(5R,7S,10R)-7-allyl-10-bromo-6-(toluene-4-sulfonyl)-6azaspiro[4.5]decan-1-one (2.36.7b).................................................................................... 171 Table 2. 21 COSY Data for (+)-(5R,7S,10R)-7-allyl-10-bromo-6-(toluene-4-sulfonyl)-6azaspiro[4.5]decan-1-one (2.36.7b).................................................................................... 172 Table 2. 22 TOCSY Data for (+)-(5R,7S,10R)-7-allyl-10-bromo-6-(toluene-4-sulfonyl)-6azaspiro[4.5]decan-1-one (2.36.7b).................................................................................... 173 Chapter 3 An Asymmetric Substituted Cyclobutanol-based Approach Towards the Synthesis of Halichlorine Table 3. 1 Attempted Oxidation of 3.18.6 .................................................................................... 218 Table 3. 2 Selective Deprotection of Bis-silyl ether 3.18.8.......................................................... 219 Table 3. 3 Oxidation of Primary Alcohol 3.18.9 .......................................................................... 220 Table 3. 4 t-Butyllithium versus Methyllithium ........................................................................... 224 Table 3. 5 Solvent Effects On the Carbonyl Addition Reaction................................................... 225 Table 3. 6 Nature of the Organometallic ...................................................................................... 226 Table 3. 7 The Effect of Temperature on the CeCl3 Promoted Carbonyl Addition Reaction ...... 227 Table 3. 8 The Effect of Solvent on the CeCl3 Promoted Carbonyl Addition Reaction in the Presence of HMPA ............................................................................................................... 228 Table 3. 9 The Effect of Using Lewis Acids to Activate Cyclobutanone 3.19.4 in the CeCl3 Promoted Carbonyl Addition Reaction ................................................................................ 228 Table 3. 10 The Effect of Adding HMPA in the CeCl3 Promoted Carbonyl Addition Reaction Where Cyclobutanone 3.19.4 is Activated with a Lewis Acid............................................. 229 Table 3. 11 Attempted Carbonyl Addition Reactions with 2-Isopropylcyclobutanone ............... 233 Table 3. 12 Formation of Alkenyl Iodide 3.22.1b........................................................................ 235 Table 3. 13 Attempted Organometallic Formation and Carbonyl Addition Reaction.................. 236 Table 3. 14 Carbonyl Addition Reaction with Alkenyl Iodide 3.22.1b ....................................... 238 Table 3. 15 Attempted Reductions of Ketone 3.23.1 ................................................................... 246 Table 3. 16 Attempted Debromination of Alcohol 3.27.4............................................................ 254  xiii Table 3. 17 Attempted Allylation of Amine 3.33.4...................................................................... 263 Table 3. 18 NMR Data for (-)-(4R,5S,7R,10S)-7-allyl-10-bromo-4-[(1S)-3-(4methoxybenzyloxy)-1-methylpropyl]-6-(toluene-4-sulfonyl)-6-azaspiro[4.5]decan-1-one (3.23.1).................................................................................................................................. 286 Table 3. 19 COSY Data for (-)-(4R,5S,7R,10S)-7-allyl-10-bromo-4-[(1S)-3-(4methoxybenzyloxy)-1-methylpropyl]-6-(toluene-4-sulfonyl)-6-azaspiro[4.5]decan-1-one (3.23.1).................................................................................................................................. 287 Table 3. 20 2D NOESY Data for (-)-(4R,5S,7R,10S)-7-allyl-10-bromo-4-[(1S)-3-(4methoxybenzyloxy)-1-methylpropyl]-6-(toluene-4-sulfonyl)-6-azaspiro[4.5]decan-1-one (3.23.1).................................................................................................................................. 288 Table 3. 21 NMR Data for (-)-(4S,5S,7R,10S)-7-allyl-10-bromo-4-isopropyl-6-(toluene-4sulfonyl)-6-azaspiro[4.5]decan-1-one (3.23.2)..................................................................... 290 Table 3. 22 COSY Data for (-)-(4S,5S,7R,10S)-7-allyl-10-bromo-4-isopropyl-6-(toluene-4sulfonyl)-6-azaspiro[4.5]decan-1-one (3.23.2)..................................................................... 291 Table 3. 23 1D Selective NOE Data for (-)-(4S,5S,7R,10S)-7-allyl-10-bromo-4-isopropyl-6(toluene-4-sulfonyl)-6-azaspiro[4.5]decan-1-one (3.23.2) ................................................... 292 Table 3. 24 NMR Data for (-)-(4R,5S,7R,10S)-7-allyl-10-bromo-4-isopropyl-6-(toluene-4sulfonyl)-6-azaspiro[4.5]decan-1-one (3.23.3)..................................................................... 294 Table 3. 25 COSY Data for (-)-(4R,5S,7R,10S)-7-allyl-10-bromo-4-isopropyl-6-(toluene-4sulfonyl)-6-azaspiro[4.5]decan-1-one (3.23.3)..................................................................... 295 Table 3. 26 1D Selective NOE Data for (-)-(4R,5S,7R,10S)-7-allyl-10-bromo-4-isopropyl-6(toluene-4-sulfonyl)-6-azaspiro[4.5]decan-1-one (3.23.3) ................................................... 296 Table 3. 27 NMR Data for Ketone 3.27.8 .................................................................................... 305 Table 3. 28 COSY Data for Ketone 3.27.8................................................................................... 306 Appendix C X-Ray Crystallography Data Table C.1 Crystallographic Data and Structure Refinement for 2.29.1........................................ 401 Table C.2 Crystallographic Data and Structure Refinement for 2.36.4a...................................... 402 Table C.3 Crystallographic Data and Structure Refinement for 2.36.7b ..................................... 403 Table C.4 Crystallographic Data and Structure Refinement for 2.37.1........................................ 404 Table C.5 Crystallographic Data and Structure Refinement for 3.25.4........................................ 405  xiv List of Figures Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids Figure 1. 1 Structures of Halichlorine, Pinnaic Acid and Tauropinnaic Acid................................ 2 Figure 1. 2 Numbering Scheme for Halichlorine ........................................................................... 3 Chapter 2 A First Generation Approach Towards Halichlorine and the Development of New Methods to form 6-Azaspirocycles Figure 2. 1 Numbering Scheme for Halichlorine ......................................................................... 36 Figure 2. 2 Representative Alkaloids Containing 6-Azaspirocenters........................................... 40 Figure 2. 3 Formation of the A Ring of Halichlorine by a Ring Closing Metathesis Reaction ... 44 Figure 2. 4 Types of Metathesis Reactions................................................................................... 46 Figure 2. 5 Common Metathesis Catalysts ................................................................................... 47 Figure 2. 6 Examples of Highly Active Second Generation Ruthenium Metathesis Catalysts.... 48 Figure 2. 7 Objectives for Bicyclic Lactam 2.1.5a....................................................................... 52 Figure 2. 8 Possible Ring Closing Metathesis Catalysts .............................................................. 54 Figure 2. 9 Second Generation Grubbs-type Catalysts................................................................. 55 Figure 2. 10 Glycal Numbering Scheme ...................................................................................... 71 Figure 2. 11 Proposed Mechanism To Explain Configuration ..................................................... 82 Figure 2. 12 NMR Data Used to Establish the Structures of 2.36.6a and 2.36.6b....................... 97 Figure 2. 13 Potential Intermediates for the Halichlorine Synthesis .......................................... 110 Figure 2. 14 ORTEP Representation of the Solid State Molecular Structure of Spirocyclopentanone 2.29.1................................................................................................ 132 Figure 2. 15 ORTEP Representation of the Solid State Molecular Structure of Ketone 2.36.4a162 Figure 2. 16 ORTEP Representation of the Solid State Molecular Structure of Spirocyclopentanone 2.36.7b ............................................................................................. 173 Figure 2. 17 ORTEP Representation of the Solid State Molecular Structure of Alcohol 2.37.1 176 Figure 2. 18 GC Trace of Spirocyclopentanones 2.41.1a and 2.41.1b....................................... 178  xv Chapter 3 An Asymmetric Substituted Cyclobutanol-based Approach Towards the Synthesis of Halichlorine Figure 3. 1 Potential Intermediate for the Halichlorine Synthesis.............................................. 202 Figure 3. 2 Steric Hindrance In the Carbonyl Addition Reaction .............................................. 204 Figure 3. 3 Will a Substituted Cyclobutanol Undergo Ring Expansion? ................................... 205 Figure 3. 4 Chiral Cyclobutanone Versus a Model Cyclobutanone ........................................... 231 Figure 3. 5 NOESY Correlations for 3.23.1 ............................................................................... 240 Figure 3. 6 NMR Data Used to Assign the Configuration of 3.23.2 .......................................... 242 Figure 3. 7 NMR Data Used to Assign the Configuration of 3.23.3 .......................................... 243 Figure 3. 8 Functionality That Would have to be Removed to Make Halichlorine ................... 243 Figure 3. 9 ORTEP Representation of Alcohol 3.25.4 ............................................................... 248 Figure 3. 10 Important 2D NMR Correlations for Ketone 3.27.8 .............................................. 256 Figure 3. 11 ORTEP Representation of the Solid State Molecular Structure of Alcohol 3.25.4 298  xvi List of Schemes Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids Scheme 1. 1 The Danishefsky Approach Towards Halichlorine...................................................... 6 Scheme 1. 2 Danishefsky’s Synthesis of Pinnaic Acid .................................................................... 7 Scheme 1. 3 Kibayashi’s Approach Towards Pinnaic Acid ............................................................. 8 Scheme 1. 4 Kibayashi’s Approach Towards Halichlorine.............................................................. 9 Scheme 1. 5 Heathcock’s Approach Towards Pinnaic Acid and Tauropinnaic Acid .................... 10 Scheme 1. 6 Heathcock’s Completion of Halichlorine .................................................................. 11 Scheme 1. 7 Uemura and Arimoto’s Approach Towards Pinnaic Acid ......................................... 12 Scheme 1. 8 Uemura and Arimoto’s Approach Towards Halichlorine.......................................... 13 Scheme 1. 9 Zhao and Ding’s Approach Towards Halichlorine and Pinnaic Acid ....................... 14 Scheme 1. 10 Martin’s Approach Towards Pinnaic Acid .............................................................. 15 Scheme 1. 11 The Martin Approach Towards Halichlorine........................................................... 16 Scheme 1. 12 Pilli’s Approach Towards Halichlorine and Pinnaic Acid....................................... 17 Scheme 1. 13 Forsyth’s Approach Towards Pinnaic Acid ............................................................. 18 Scheme 1. 14 The Wright Approach Towards Halichlorine .......................................................... 18 Scheme 1. 15 Clive’s First Generation Approach Towards Halichlorine ...................................... 19 Scheme 1. 16 Clive’s Second Generation Approach Towards Halichlorine.................................. 20 Scheme 1. 17 Clive’s Third Generation Approach Towards Halichlorine..................................... 21 Scheme 1. 18 The Simpkins First Approach Towards Halichlorine and Pinnaic Acid.................. 22 Scheme 1. 19 The Simpkins Second Approach Towards Halichlorine and Pinnaic Acid ............. 23 Scheme 1. 20 White’s Approach Towards Halichlorine ................................................................ 23 Scheme 1. 21 The Feldman Approach Towards Halichlorine........................................................ 24 Scheme 1. 22 The Ihara Approach Towards Halichlorine and Pinnaic Acid................................. 25 Scheme 1. 23 Mol and Bubnov’s Approach Towards Halichlorine and Pinnaic Acid .................. 26 Scheme 1. 24 The Shishido and Itoh First Generation Approach Towards Halichlorine .............. 27 Scheme 1. 25 The Shishido and Itoh Second Generation Approach Towards Halichlorine.......... 28 Scheme 1. 26 Zhao’s Approach Towards Halichlorine.................................................................. 29 Scheme 1. 27 The Stockman Approach Towards Halichlorine...................................................... 30 Scheme 1. 28 Weinreb’s Synthesis of the C15-C21 Subunit of Halichlorine and Pinnaic Acid ...... 31 Scheme 1. 29 Contributions from the Taber Group ....................................................................... 31  xvii Chapter 2 A First Generation Approach Towards Halichlorine and the Development of New Methods to form 6-Azaspirocycles Scheme 2. 1 Retrosynthetic Analysis for Halichlorine................................................................. 37 Scheme 2. 2 The Pinacol Rearrangement ..................................................................................... 38 Scheme 2. 3 Acid Catalyzed Semipinacol Rearrangement to Form 6-Azaspirocenters............... 40 Scheme 2. 4 Proposed Mechanistic Rationale for the Observed Diastereoselectivity in the Acid Mediated Semipinacol Rearrangement of 2.3.3b ................................................................. 42 Scheme 2. 5 The Siloxy-epoxide Semipinacol Rearrangement Approach Towards Fasicularin . 43 Scheme 2. 6 Synthesis of Dienes 2.1.6a and 2.1.6b..................................................................... 45 Scheme 2. 7 Chauvin’s Proposed Mechanism for Metal Catalyzed Alkene Metathesis.............. 50 Scheme 2. 8 Proposed Mechanism for Ruthenium Metathesis Catalysts..................................... 51 Scheme 2. 9 Attempted Elaboration of 2.1.5a.............................................................................. 53 Scheme 2. 10 Synthesis of Diphenyl Propargyl Alcohol 2.10.4 .................................................. 56 Scheme 2. 11 Synthesis of the N-Heterocyclic Carbene Ligand .................................................. 56 Scheme 2. 12 Synthesis of the Second Generation Catalyst 2.9.3 ............................................... 57 Scheme 2. 13 Ring Closing Metathesis and Attempted Elaboration of 2.1.6b ............................ 58 Scheme 2. 14 Retrosynthetic Analysis for Halichlorine via Path B ............................................. 59 Scheme 2. 15 Synthesis of Alkenyl Stannane 2.1.11 ................................................................... 60 Scheme 2. 16 Synthesis and Attempted Acid Catalyzed Semipinacol Rearrangement of 2.1.10 62 Scheme 2. 17 Two Possible Mechanisms for the Addition of Halogens (X2) to Alkenes ........... 66 Scheme 2. 18 The Addition of Bromine and Water to 1-Methyl-1-cylopentene ......................... 68 Scheme 2. 19 Addition Reactions with Substituted Cyclohexene Compounds ........................... 69 Scheme 2. 20 Reactions of Cyclic Allylic Alcohols/Ethers ......................................................... 70 Scheme 2. 21 Reactions of Cyclic Allylic Alcohols..................................................................... 70 Scheme 2. 22 Reactions of Cyclic Allylic Alcohols with an Internal Nucleophile...................... 71 Scheme 2. 23 Possible Transition States for the Reaction of Glycals with Electrophilic Halogen Reagents................................................................................................................................ 72 Scheme 2. 24 The First Electrophilic Halogen Promoted Ring Expansion Reactions................. 76 Scheme 2. 25 Ring Expansions of Cyclopropanols...................................................................... 76 Scheme 2. 26 Ring Expansion Reactions with Bromine and tert-Butyl Hydroperoxide ............. 77 Scheme 2. 27 Synthesis of Chiral Bromo-Camphor Derivatives ................................................. 78 Scheme 2. 28 Paquette’s NBS Promoted Ring Expansion via an Oxonium Ion.......................... 79 Scheme 2. 29 Successful Semipinacol Rearrangement Promoted by N-Bromosuccinimide ....... 80  xviii Scheme 2. 30 Potential Ring Expansion Substrates ..................................................................... 84 Scheme 2. 31 Retrosynthetic Analysis of the Semipinacol Rearrangement Substrates ............... 84 Scheme 2. 32 Synthesis of 4-Substituted Amides 2.32.3, 2.32.4 and 2.32.5 ............................... 86 Scheme 2. 33 Synthesis of 5-Substituted Amide 2.30.5............................................................... 86 Scheme 2. 34 Synthesis of 6-Allyl Amides 2.30.7 and 2.30.8 ..................................................... 87 Scheme 2. 35 Coupling of Enol Triflates with Gilman Reagents................................................. 89 Scheme 2. 36 Synthesis of 5-Benzyloxy Substrate 2.3.3a ........................................................... 92 Scheme 2. 37 Ring expansion and ORTEP of Compound 2.36.4a .............................................. 95 Scheme 2. 38 Confirmation of Configuration for 2.36.7b ......................................................... 100 Scheme 2. 39 Proposed Mechanism for the Ring Expansion of 4-Substituted Substrates......... 101 Scheme 2. 40 Proposed Mechanism for the Ring Expansion of the 5-Substituted Substrates... 102 Scheme 2. 41 Proposed Mechanism for the Ring Expansion of the 6-Substituted Substrates... 103 Scheme 2. 42 Assumed Product Outcome for the Acid Catalyzed Rearrangement of Substrates with Substituents at the 4-position...................................................................................... 106 Scheme 2. 43 Stereochemical Assignments for Compounds 2.3.4d, 2.3.4e and 2.3.5e............. 107 Scheme 2. 44 Model Used to Predict the Configuration of the Product Formed in the Acid Catalyzed Semipinacol Rearrangement of Allylic Alcohol 2.30.6..................................... 108 Scheme 2. 45 Chemical Correlation of 2.36.7b and 2.41.2b...................................................... 109 Scheme 2. 46 Attempted Elaboration of Compound 2.29.1 ....................................................... 110 Scheme 2. 47 Siloxy-epoxide Semipinacol Rearrangements ..................................................... 111 Scheme 2. 48 Siloxy-Epoxide Rearrangement of 2.30.6............................................................ 113 Scheme 2. 49 Chemical Correlation of Compound 2.48.3 with Compound 2.36.7b................. 114 Chapter 3 An Asymmetric Substituted Cyclobutanol-based Approach Towards the Synthesis of Halichlorine Scheme 3. 1 New Retrosynthetic Analysis for Halichlorine ........................................................ 203 Scheme 3. 2 Alkylation of the Hydrazone of Cyclobutanone ...................................................... 206 Scheme 3. 3 Bergman’s Synthesis of 2-Substituted Cyclobutanones from Cobaltocyclopentanones ....................................................................................................... 207 Scheme 3. 4 Stryker’s Synthesis of Substituted Cyclobutanones................................................. 208 Scheme 3. 5 Van Leusen’s TosMic Approach Towards Substituted Cyclobutanones................. 208 Scheme 3. 6 Substituted Cyclobutanones via [2+2] Cycloaddition Reactions............................. 209  xix Scheme 3. 7 Synthesis of 2-Alkoxycyclobutanones..................................................................... 209 Scheme 3. 8 Synthesis of 2-Alkenyl Cyclobutanones .................................................................. 210 Scheme 3. 9 Trost’s Synthesis of Substituted Cyclobutanones From Cyclopropylphenylsulfide 211 Scheme 3. 10 Gadwood’s Synthesis of Substituted Cyclobutanones........................................... 211 Scheme 3. 11 Synthesis of Cyclobutanones via Epoxidation or Dihydroxylation of Cyclopropylidenes ................................................................................................................ 212 Scheme 3. 12 Kulinkovich Cyclopropanation Approach Towards Substituted Cyclobutanones 212 Scheme 3. 13 Bergman’s Synthesis of Optically Active Cyclobutanones ................................... 213 Scheme 3. 14 Salaün’s Syntheses of Optically Active Cyclobutanones ...................................... 214 Scheme 3. 15 Asymmetric Epoxidation Route Towards Optically Active Cyclobutanones ....... 215 Scheme 3. 16 Asymmetric Dihydroxylation Route Towards Optically Active Cyclobutanones. 215 Scheme 3. 17 Cha’s Route Approach Towards Optically Active Cyclobutanones...................... 216 Scheme 3. 18 Synthesis of Chiral Epoxide 3.18.5........................................................................ 217 Scheme 3. 19 Completion of the Synthesis of Chiral Cyclobutanone 3.19.4............................... 220 Scheme 3. 20 Conformational Analysis to Rationalize the Formation of Cyclobutanone 3.19.4 221 Scheme 3. 21 Synthesis of 2-Isopropylcyclobutanone 3.20.1 ...................................................... 231 Scheme 3. 22 Can the Organometallic Be Made from an Alkenyl Halide? ................................. 234 Scheme 3. 23 N-Bromosuccinimide Promoted Ring Expansion Reactions of Substituted Cyclobutanols ....................................................................................................................... 238 Scheme 3. 24 Radical Formation of 3.23.1 .................................................................................. 245 Scheme 3. 25 Mechanism for the Formation of Vinylogous Amide 3.24.2................................. 246 Scheme 3. 26 Radical Cyclization of Bromide 3.25.4.................................................................. 249 Scheme 3. 27 Ozonolysis and Acetal Formation Prior to Radical Cleavage of the Bromine ...... 251 Scheme 3. 28 Proposed Mechanism for the Formation of Ketone 3.27.8 .................................... 257 Scheme 3. 29 Explanation for the Formation of epi-3.27.8.......................................................... 257 Scheme 3. 30 E2 Elimination and Reduction of Bromide 2.36.8a ............................................... 259 Scheme 3. 31 E2 Elimination and Subsequent Test Reactions of Alcohol 3.30.2........................ 259 Scheme 3. 32 E2 Elimination and Reduction of the Ketone of 3.23.1 .......................................... 260 Scheme 3. 33 Formation of Bis Silyl ether 3.33.2........................................................................ 261 Scheme 3. 34 Deprotection and Allylation of Bis Silyl Ether 3.33.4........................................... 264 Scheme 3. 35 Summary of the Synthesis of Chiral Non-racemic Cyclobutanone 3.19.4 ............ 265 Scheme 3. 36 Summary of Synthesis of Acrylate 3.34.1 ............................................................. 266  xx List of Symbols and Abbreviations Å A1,3 strain α Ac acac ADMET AIBN amu anal. APT Ar atm ax β 9-BBN Bn Boc b.p. br brsm Bu Bz °C calcd concd COSY CI CM CR CSA cPLA2 Cn Cy d dd δ δ+ ∆ d 2D dba DBU DCI DDQ DEAD DPPA dppf DIBAL-H  -  -  angstrom allylic 1,3-strain below the plane of a ring acetyl acetylacetonate acyclic-diene metathesis polymerization 2,2’-azobisisobutyronitrile atomic mass unit analysis attached proton test aryl atmosphere axial above the plane of a ring 9-borabicyclo[3.3.1]nonane benzyl tert-butyloxycarbonyl boiling point broad based on recovered starting material butyl benzoyl degrees Celsius calculated concentrated (1H-1H) homonuclear correlation spectroscopy chemical ionization cross metathesis Comins’ reagent (1S)-(+)-10-camphorsulfonic acid cytosolic phospholipase carbon number n where n = 1,2,3… cyclohexyl doublet doublet of doublets chemical shift partial positive charge heat days two-dimensional dibenzylideneacetone 1,8-diazabicyclo[5.4.0]undec-7-ene desorption chemical ionization 2,3-dichloro-5,6-dicyano-1,4-benzoquinone diethyl azodicarboxylate diphenylphosphoryl azide bis(diphenylphosphino)ferrocene diisobutylaluminum hydride  xxi (-)-DIT DMA DMAP DMDO DME DMF DMP DMPU dr DMSO E E E2 ed. Ed., Eds. EDCI ee EI endo epi Eq. eq equiv ESI Et Et2O exo g gem GC h HCl HF·pyr HFIP HMBC HMPA HMQC HOBT HOMO HPLC hυ HRMS Hn  -  (-)-diisopropyl tartrate N,N-dimethylacetamide 4-dimethylaminopyridine dimethyldioxirane 1,2-dimethoxyethane N,N-dimethylformamide Dess Martin periodinane N,N'-dimethyl-N,N'-propylene urea diastereomeric ratio dimethyl sulfoxide electrophile entgegen (configuration) elimination bimolecular edition editor, editors 3-(3-dimethylaminopropyl)-N-ethylcarbodiimide enantiomeric excess electron ionization endocyclic epimeric equation equatorial equivalent electrospray ionization ethyl diethyl ether exocyclic gram(s) geminal gas chromatography hour(s) hydrochloric acid hydrogen fluoride-pyridine complex 1,1,1,3,3,3-hexafluoroisopropyl heteronuclear multiple bond coherence hexamethylphosphoramide heteronuclear multiple quantum coherence 1-hydroxybenzotriazole highest occupied molecular orbital high pressure liquid chromatography light high resolution mass spectrometry hydrogen number n where n = 1,2,3…  Hz i IC50  -  Ipc i Pr  -  hertz (s ) iso the molar concentration of an antagonist that produces 50% of the maximum inhibitory response for that antagonist isopinocampheyl isopropyl  -1  xxii i.e. Im IR IUPAC J JSn-H kg KHMDS LA LDA LG LHMDS LUMO m M M+ MCPBA or m-CPBA Me Mes MeCN min mg MHz µL mmol MOM m.p. Ms MS n NBS ND NIS NHC NMO NMR NOE NOESY Nuc [O] OTf ORTEP p pdt PG Ph PhCH3 PhH PMB  -  Id Est (that is) imidazole infrared International Union of Pure and Applied Chemistry coupling constant coupling constant for tin and proton nuclei kilogram(s) potassium bis(trimethylsilyl)amide Lewis acid lithium diisopropylamide leaving group lithium bis(trimethylsilyl)amide lowest unoccupied molecular orbital multiplet molar molecular ion meta-chloroperoxybenzoic acid methyl mesityl or 2,4,6-trimethylphenyl acetonitrile minute(s) miligram(s) megahertz microlitre(s) milimole(s) methoxymethyl melting point methanesulfonyl (mesyl) molecular sieves normal N-bromosuccinimide not determined N-iodosuccinimide N-heterocyclic carbene N-methylmorpholine-N-oxide nuclear magnetic resonance nuclear Overhauser enhancement nuclear Overhauser enhancement spectroscopy nucleophile oxidizing conditions trifluoromethylsulfonate (triflate) Oak Ridge Thermal Ellipse Program para product protecting group phenyl toluene benzene para-methoxybenzyl  xxiii PNB ppm PPTS Pr psi pyr Q q quin rt R R RCM rec’d ROM ROMP S s SM SN1 SN2 T t t-Bu t-BuOK t-BuOMe tert TANCA TBAF TBAI TBS TES Tf TFA THF TIPS TLC TMS Tol TosMic Tosyl TPAP trig Ts TsCl TsOH μg μmol VCAM-1  -  para-nitrobenzoate parts per million pyridinium para-toluenesulfonate propyl pounds per square inch pyridine quartenary quartet quintet room temperature alkyl group, aryl group or hydrogen atom rectus (configuration) ring-closing metathesis recovered ring opening metathesis ring opening metathesis polymerization sinister (configuration) singlet starting material substitution nucleophilic unimolecular substitution nucleophilic bimolecular temperature time tert-butyl potassium tert-butoxide tert-butylmethyl ether tertiary transannular nitrone cycloaddition tetrabutylammonium fluoride tetrabutylammonium iodide tert-butyldimethylsilyl triethylsilyl trifluoromethanesulfonyl trifluoroacetic acid tetrahydrofuran triisopropylsilyl thin layer chromatography trimethylsilyl para-tolyl para-toluensulfonylmethylisocyanide para-toluenesulfonyl or toluene-4-sulfonyl tetra-n-propylammonium perruthenate trigonal para-toluenesulfonyl (tosyl) para-toluenesulfonyl chloride para-toluenesulfonic acid microgram micromole vascular cell adhesion molecule-1  xxiv X yld Z 18-C-6 · ±  -  generic halogen (i.e. F, Cl, Br or I) yield zusammen (configuration) 18-crown-6 coordination complex racemic  ©  -  registered trademark copyright  ®  xxv Acknowledgements I would like to thank my supervisor Professor Gregory Dake for his mentorship and support over the years. I appreciate all the time he has given to me and I am grateful for all that I have learned from him. I especially enjoyed our group meetings and the many conversations that we have shared. I would like to thank Professor Dake and Professor Marco Ciufolini for their help in proof reading this thesis. Any and all errors of a scientific nature or misuse of the English language are my own. I want to acknowledge the staff of the NMR room especially Nick Burlinson for his help running experiments and with interpreting spectra. I also want to thank Brian Patrick for helping me to prove various stereochemical issues with several of my compounds. Any help provided to me by the mass spectrometry laboratory is gratefully acknowledged. Financial assistance, specifically provided by the C. A. McDowell entrance scholarship and a Gladys Estella Laird fellowship, was particularly appreciated. I owe my choice in career path to Professor Graham Bodwell. Before I took your 3rd year organic chemistry class I didn’t particularly enjoy organic chemistry. You challenged me in a positive way and you showed me how much fun synthetic organic chemistry could be. Without you who knows what career path I would have chosen instead. Thank you. I would like to thank all of the Dake group members, Leah, Tyler, Wilson, Michaël, Erik, Melissa, Jacqueline, Jen, Krystle and Julien, in addition to the various summer and 449 students, for making my time in the lab enjoyable. If you guys weren’t around I probably would have went crazy long ago. A special thanks is given to Leah, Tyler, Noonan, Wayne, Jay, Richard and Emily. You guys have been special friends to me and you have been there through both the low moments and through many of the good times. I have cherished our time together (although I am not sure I can speak for my liver) and I hope that we will remain friends long after I have left UBC. I want to give a special thankyou to my parents, Nick and Sylvia. You taught me the value of education at a very young age. Through the years you have loved me and supported me and for this I will be forever thankful. I also want to say thankyou to my siblings Natasha, Andrea and Kenneth. You mean more to me than I ever let you know and I appreciate all the help you have given to me, especially during the writing of this thesis.  xxvi Foreword The field of synthetic organic chemistry can be sub-divided into two closely related areas: target-oriented synthesis and methodology. Target-oriented synthesis involves the selection of a target molecule and the development and implementation of a synthetic plan to synthesize that molecule. Methodology can be described as the development of reaction methods to carry out a specific transformation. Quite often there is a symbiotic relationship between the two subclasses of synthetic organic chemistry. Sometimes the synthesis of certain compounds will require a transformation to be done for which there is no existing method or for which existing methodologies are inefficient. This often leads to the development of new reaction methods. Sometimes new methodologies are developed which someone later figures out can be used to synthesize a specific molecule. There are many other reasons why people do organic synthesis. Sometimes synthesis is required to confirm or to establish the configurations found in complex natural products. In other instances molecules are synthesized for their therapeutic or commercial value. Other people synthesize compounds because of the inherit challenges that the molecules possess. Finally, some people synthesize compounds to elucidate reaction mechanisms or to figure out how different factors affect certain reactions. In the following thesis, a synthetic approach towards the synthesis of the natural product halichlorine is presented. Initially a discussion regarding the previous synthetic routes towards halichlorine is given. This is followed by the presentation of a first generation approach towards halichlorine that is then followed by a second generation approach towards halichlorine. The eventual synthetic route was discovered through a sequence of success, failure and problem solving. The research led to the discovery of new reaction methods that resulted in the synthesis of a complex intermediate towards the total synthesis of halichlorine. A couple of comments are given here regarding the numbering system used to identify compounds throughout this thesis. First, the compounds presented in a specific scheme are presented chronologically and are meant to reflect the both the chapter number and the scheme number. For example, Scheme 2.30 is the thirtieth scheme found in chapter 2. The first new compound presented in Scheme 2.30 is labeled 2.30.1. The second new compound in Scheme 2.30 is labeled 2.30.2 and so on. Labels are applied only once and therefore the same label is used throughout the remainder of the thesis. This means that compound 2.30.2 will be labeled 2.30.2 even if it appears in chapter 3. Compounds are only labeled according to scheme number.  xxvii New compounds that appear in equations, figures and tables are not labeled to match the equation, figure or table number. These compounds are labeled in sequence following the label that was assigned to the last compound to appear in the preceding scheme. For example, compound 2.30.4 is the last compound to appear in Scheme 2.30. The next diagram, Equation 2.12 contains the next new compound in chapter 2, this new compound is labeled 2.30.5. This numbering scheme will hopefully make it easier to find compounds because schemes should be easier to find than individual compounds. The reader should be aware that, with the exception of chapter 1, this thesis is written in a stream of consciousness format. Generally speaking I will present an idea, I will discuss previous work related to that idea, I will present the work that I did and the results will be discussed. The results are then used to determine the nature of the work that follows. This process is repeated throughout the thesis. In certain instances I felt it was necessary to elaborate on a certain topic and therefore the reader should expect to encounter sections where special topics are discussed. Finally, at the end of each chapter, the reader will find a summary of the important contributions presented in that chapter.  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1 Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1  2  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.1  Introduction As part of a program to find new biologically active compounds, halichlorine (I) was  isolated from the marine sponge Halichondria okadai Kadota by Uemura and co-workers in 1996 (Figure 1. 1). 1 Subsequent work by the Uemura group resulted in the isolation of the structurally related compounds pinnaic acid (II) and tauropinnaic acid (III) from the Okinawan bivalve Pinna muricata (Note: the structures for pinnaic acid (II) and tauropinnaic acid (III) depicted below are arbitrarily represented in their zwitterionic forms). The absolute configuration of these molecules was determined by extensive spectroscopic experiments, partial degradation and/or through chemical synthesis.1,2 These molecules exhibit interesting biological activity (see below) and they present formidable targets to the synthetic community. Figure 1. 1 Structures of Halichlorine, Pinnaic Acid and Tauropinnaic Acid H  H O O  Cl  HO  H  HO  H2N O  Cl  I  1.2  +  O  N  -  H  O O S O -  H N  H  II  H2N  HO  H  Cl  HO  +  O  HO III  Biological Activity Halichlorine (I) was found to inhibit vascular cell adhesion molecule-1 (VCAM-1) with  an IC50 of 7μg/mL. VCAM-1, a member of the immunoglobulin family is expressed on the surface of endothelium cells. The interaction between endothelium cells and leukocytes involves VCAM-1. It also participates in the migration of leukocytes to inflammatory foci. This essentially means that there is a direct link between VCAM-1 and inflammatory diseases. VCAM-1 has been found to play a role in rheumatoid arthritis, allergies and nephrotoxic nephritis. In addition the adhesive nature of VCAM-1 is used by cancer cells in the metastatic process. For the preceding reasons a VCAM-1 inhibitor may be a useful treatment for some inflammatory diseases and some cancers. Pinnaic acid (II) and tauropinnaic acid (III) were found to inhibit a cytosolic phospolipase cPLA2 with an IC50 of 0.2 mM for II and an IC50 of 0.09 mM for III. Inhibitors for  3  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  cPLA2 are also potentially useful for the treatment of inflammatory diseases. This is due to the fact that cPLA2 is involved at an early stage in the series of reactions that leads to the formation of inflammatory mediators: leukotrienes, lipoxins, prostaglandins and thromboxanes. 1.3  Synthetic Analysis Halichlorine and the pinnaic acids are challenging synthetic targets. Consequently they  have garnered much attention from various research groups. Derrick Clive and co-workers have documented this work in an excellent review. 3,4 While it is unnecessary to give a detailed account of this work an abbreviated discussion will be presented. In examining the various approaches to these molecules only the “important transformations” will be discussed. “Important transformations” will be defined as those where a stereocenter is established, where a ring is formed or where a significant portion of the molecule is introduced. Before the various contributions are presented an analysis of the architecture of halichlorine and the pinnaic acids should be given. After identifying the specific challenges presented by these molecules the approaches taken by each research group will be given with an emphasis on how each group dealt with the particular synthetic problems. Figure 1. 2 Numbering Scheme for Halichlorine H A  O O  Cl  N  D H  O O S O  H B  +  H2N  O C  HO  O  Cl  HO  I  -  H N  H +  H2N  O HO  H  H  Cl  HO II  HO III  4  H  3 5  O  HN  21  [4.5] 6-azaspirodecane  1 2  N H  O  9  8 10  14 13 12  11  17  Cl  HO  Halichlorine arguably has the most complex structure of the three natural products in that it has more rings (4) than either pinnaic acid or tauropinnaic acid. The [4.5] 6-azaspirodecane ring system is common to all three natural products (Figure 1. 2) and represents one of the key  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  4  synthetic challenges. The numbering system used to describe the [4.5] 6-azaspirodecane ring system is in accord with IUPAC nomenclature for these types of ring systems. The numbering system used to describe halichlorine in Figure 1. 2 is based on the numbering system applied to the natural product by Uemura and co-workers following its isolation. All of the groups that have published work involving the synthesis of halichlorine have adopted this numbering system and therefore this numbering system will be used to describe halichlorine throughout this document. The [4.5]6-azaspirodecane ring system will be referred to as the BC ring system. Halichlorine differs from the pinnaic acids in that it also has a 6-membered ring fused to the B ring which will be referred to as the A ring. As well, halichlorine has a fourth ring, a 15-membered macrolactone that will be referred to as the D ring. Otherwise all three natural products have similar carbon skeletons and identical stereochemical configurations. There are five stereogenic centers present in the molecules: the C5 secondary carbon adjacent to the nitrogen, the C9 tertiary spirocenter adjacent to the nitrogen, the C13 tertiary carbon adjacent to the spirocenter, the C14 secondary carbon with the only methyl substituent and finally the C17 secondary alcohol. There are three alkenes: two trisubstituted alkenes, one of which has a chlorine substituent, and a disubstituted alkene. Almost all synthetic approaches towards these natural products have involved the assembly of the [4.5] 6-azaspirodecane ring system. Some groups have synthesized the ABC tricyclic ring system while others have synthesized all three natural products. Finally a couple of groups have proposed ways of introducing the C14 to C21 segment. In the following section a discussion of the approaches towards halichlorine and the pinnaic acids will be presented with attention to how the synthetic problems were addressed. 1.4 1.4.1  Synthetic Approaches Towards Halichlorine and the Pinnaic Acids Introduction As mentioned various research groups have published synthetic approaches towards  halichlorine and the pinnaic acids, a summary of this work follows. The work done on the tricyclic core will be presented first followed by work done on the C14 to C21 segment. The work on the tricyclic core is divided into three sections: 1) groups that start with the C ring, 2) groups that start with the B ring and finally 3) groups that make both the B and C rings in one step.  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  5  Within each section the work will be presented by research group and with some exceptions the work will generally be presented in order of importance. This means that groups that have completed either total or formal syntheses will be presented first. Groups that have published uncompleted approaches or groups that have studied model compounds will follow. Whenever groups have published similar strategies an attempt will be made to present these strategies together. 1.4.2  Groups that Start with the C Ring  1.4.2.1 Contributions from the Danishefsky Laboratory: Total Synthesis of Halichlorine2b,c,d The Danishefsky group reported the first asymmetric total synthesis of halichlorine starting with the known Meyers lactam 1.1.1 (Scheme 1. 1). As depicted in Scheme 1.1 the C ring of Meyers lactam would become the C ring of halichlorine and one of the bridgehead carbons has the desired C13 configuration. Allylation on the convex face of the bicyclic system established what would become the C9 spirocenter. Following a series of transformations, alkylation with methyl iodide introduced the C14 methyl group with the desired configuration. Straightforward functional group manipulations produced cyclopentyl amine 1.1.5. Treatment of this substrate with acid resulted in deprotection of the tert-butoxycarbonyl (Boc) protecting group to give the free 1° amine, which when neutralized with potassium carbonate, underwent spontaneous intramolecular Michael addition with the enoate moiety to produce 6-azaspirocycle 1.1.7. The configuration of the product is likely the result of the reaction proceeding through conformation 1.1.6 where the larger substituent is situated pseudoequatorially in a chair-like transition state. This reaction closed the B ring and set the C5 configuration. After β-keto ester 1.1.8 was made the A ring of the tricyclic core was formed by a Mannich reaction with formaldehyde resulting in an inconsequential mixture of diastereomers and tautomers. Tricycle 1.1.9 was elaborated to alkyne 1.1.10, which when treated with the Schwartz reagent (Cp2Zr(H)Cl) in the presence of dimethyl zinc produced an intermediate alkenyl zinc species 1.1.11. Addition of this alkenyl zinc compound to enal 1.1.12 in the presence of chiral aminoalcohol 1.1.13 gave allylic alcohol 1.1.14 as a 4:1 mixture of epimers in favour of the one shown. This reaction introduced the remainder of the C14-C21 segment and established the C17 configuration. After some protecting group manipulations, macrolactonization of 1.1.15 closed  6  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  the D ring, which after deprotection of the silyl ether gave halichlorine (I). Scheme 1. 1 The Danishefsky Approach Towards Halichlorine Ph O  N O  13  N  TiCl4  C  -78 oC to rt (99 %)  H 1.1.1  HO  Ph  TMS  O  Boc N  9  O  H 1.1.2  MeO2C  H 1.1.3  i) (Me3Si)2NLi  Boc N  o  THF, -40 C  O  ii) MeI, -78 oC to 0 oC (90 %)  14  H 1.1.4  OSiPh 2t-Bu BocHN t-BuPh2SiO  MeO2C  i) CF3CO2H, CH2Cl2  H 1.1.6  H  H  H  O  MeO2C  5  HN  B  O  CH2O, EtOH (73 %)  HN  t-BuO 2C  t-BuPh2SiO 1.1.7  H  N  ii) aq K2CO3 (77 % over 2 steps)  H 1.1.5  H  A t-BuO 2C  t-BuPh2SiO  H  N  t-BuPh2SiO 1.1.8  H  H 1.1.9 OSiMe 2t-Bu  H  t-BuO 2C  i)Cp2Zr(H)Cl CH2Cl2  N  Cl O 1.1.12 Ph  N  t-BuO 2C  ii) Me2Zn o  heptane, -65 C H 1.1.10  N Me  H  MeZn 1.1.11  OSiMe 2t-Bu  Ph  H  OH  17  Cl  EDCI, DMAP HCl, THF, heat (54 %)  OH  HO 1.1.14  H  N  O  N O D  N  t-BuO 2C  1.1.13  H  HO2C  H  H  H  H O  H  N O  H  H Cl  Cl OSiMe 2t-Bu  OSiMe 2t-Bu  1.1.14  1.1.15  Cl  HO halichlorine  1.4.2.2 Contributions from the Danishefsky Laboratory: Synthesis of Pinnaic Acid2b,c,d The Danishefsky group also synthesized pinnaic acid (II) starting from bicyclic lactam 1.1.4 (Scheme 1. 2), which was also used in their synthesis of halichlorine (I). The bicyclic lactam was elaborated to give cyclopentyl amine 1.2.1. Hydroboration of the terminal alcohol, followed by Suzuki coupling with vinyl iodide 1.2.2 gave conjugated diene 1.2.3 in good yield.  7  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  Deprotection of the nitrogen and subsequent base-promoted cyclization gave 6-azaspirocycle 1.2.4. Similar to the halichlorine synthesis this sequence closed the B ring and established the C5 configuration. This compound was easily converted to aldehyde 1.2.5. The Horner-WadsworthEmmons reaction with phosphonate 1.2.6 was a difficult transformation as the reaction failed to go to completion and the unreacted aldehyde 1.2.5 was inseparable from the desired product 1.2.7. Interestingly, reduction of the mixture with either enantiomer of Alpine hydride gave the same diastereomeric alcohol 1.2.8 in addition to the alcohol produced from the reduction of aldehyde 1.2.5. This alcohol byproduct could be separated, reoxidized to aldehyde 1.2.5 and recycled through the synthetic sequence again. The desired allylic alcohol 1.2.8 could be obtained in ~30 % yield from aldehyde 1.2.5. Overall, this process established the C17 configuration. Subsequent removal of the protecting groups and ester hydrolysis gave pinnaic acid (II). Scheme 1. 2 Danishefsky’s Synthesis of Pinnaic Acid 9-BBN, THF Pd(dppf)2Cl2  Boc NH  Boc N O t-BuPh2SiO  H 1.1.4  H  EtO 2C  1.2.1 H EtO 2C  1) CF3CO2H, CH2Cl2 2) DBU (81 % 2 steps)  5  HN t-BuPh2SiO  THF-HMPA, -78 C to rt, 2 days; t-BuMe2SiO  Cl O 1.2.6 EtO 2C t-BuMe2SiO  Cl  F 3CCO 17  PO(OMe)2  t-BuPh2SiO  1.2.2  H  F 3CCO O  H 1.2.5  H  F 3CCO  (R) or (S)-Alpine hydride  N  H Cl 1.2.7  H  H HO2C HO  HN  deprotections and ester hydrolysis H  HO 1.2.8 [~ 30 % from 1.2.5 plus reduced 1.2.5 (~ 50 %)]  N  H  O  N  1.2.3  EtO 2C  EtO 2C t-BuMe2SiO  o  BocHN I  B  1.2.4 (Me3Si)2NLi, THF, -78 oC  EtO 2C  Ph3As, Cs2CO3 THF-water (75 %)  H Cl HO pinnaic acid (II)  H  8  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.4.2.3 The Kibayashi Approach Towards Pinnaic Acid  5  The Kibayashi approach to pinnaic acid (II) began with hydroxamic acid 1.3.1, which contains the C ring of halichlorine (Scheme 1. 3). Oxidation with tetrapropylammonium periodate generated an acyl nitroso species 1.3.2 that underwent a spontaneous ene reaction to produce 6-azaspirocycle 1.3.3. This reaction formed the B ring and set the C13 stereocenter. After straightforward manipulations addition of lithium acetylide to the lactam carbonyl of 1.3.4, followed by reduction of an in situ generated iminium ion (not shown) gave alkyne 1.3.5 as the only product and thus established the C5 stereocenter. Following a series of steps, tricyclic alkene 1.3.6 was hydrogenated stereoselectively to give 1.3.7, which has the desired configuration for C13. After cleavage of the silylether, alkylation with methyl iodide gave predominantly (~ 15:1) the desired configuration at C14. Through a series of transformations, this compound was converted into aldehyde 1.3.9 that was subsequently olefinated with phosphonate 1.3.10 to give α,β-unsaturated ester 1.3.11. This compound is the racemic version of a compound used in Danishefsky’s synthesis of pinnaic acid (II). Scheme 1. 3 Kibayashi’s Approach Towards Pinnaic Acid HOHN  O  MOMO  O Pr4NIO4, CHCl3 0 oC (82 %)  C  O  OMOM  1.3.1  MOMO  1.3.2  ii) NaBH3CN, AcOH (67 %)  B  HO N  H  i) H Li H2NCH2CH2NH2, THF, 5 oC  O  O  N  BnO  1.3.3  1.3.4  t-BuPh2SiO  1) H2, Pd/C, EtOH (99 %) 2) 1M HCl, THF, (99 %)  5  BnO  N  O  N  MOMO 1.3.5  1.3.6  H  HO  HO O  N 13  H  LDA, MeI, THF - 78 oC (78 %) 15:1 ratio of epimers  O O  N 14  H  1.3.7  F 3CCO t-BuPh2SiO  1.3.8  1.3.9  H EtO 2C PO(OEt) 2 1.3.10 NaH, THF, -78 oC (98 %)  N  MOMO  9  EtO 2C F 3CCO  N  t-BuPh2SiO 1.3.11  Pinnaic Acid (II) H  N  H  9  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.4.2.4 Kibayashi’s Completion of Halichlorine  5  Using aldehyde 1.3.9 from their pinnaic acid (II) synthesis, the Kibayashi group also completed a formal synthesis of halichlorine (I) (Scheme 1. 4). Aldehyde 1.3.9 was transformed into 6-azaspirocycle 1.4.1 in a couple of steps. Alkylation of the nitrogen was done with activated allylbromide 1.4.2. Subsequent treatment with a second generation Grubbs catalyst effected a clean intramolecular metathesis reaction which closed the A ring and formed tricycle 1.4.4. Removal of the silyl protecting group gave alcohol 1.4.5; the Danishefsky group used the corresponding optically pure tert-butyl ester in their synthesis of halichlorine (I). Scheme 1. 4 Kibayashi’s Approach Towards Halichlorine H  H  H  O F 3CCO  N  t-BuPh2SiO  EtO 2C 1.4.2  HN t-BuPh2SiO  H 1.3.9  Grubbs II, reflux, CH2Cl2 (99 %)  A EtO 2C  N  t-BuPh2SiO  H  MeCN, K2CO3, 60 oC (88 %)  1.4.1  H  Br N  EtO 2C t-BuPh2SiO  H  1.4.3  H N  EtO 2C  halichlorine (I)  HO H 1.4.4  H 1.4.5  1.4.2.5 Heathcock’s Synthesis of Pinnaic Acid and Tauropinnaic Acid2e The Heathcock group has synthesized the three members of this family of natural products in racemic form.2e Reduction of the known enamine ester 1.5.1 provided bicycle 1.5.2 (Scheme 1. 5). This compound contains the C ring and has both the C13 and C14 stereocenters set. After the pyrrolidine function was converted to a benzyl carbamate, stereoselective allylation gave the highly functionalized cyclopentane 1.5.4. This reaction established the C9 configuration. This compound was elaborated to the α,β-unsaturated ketone 1.5.5 that when treated with hydrogen and palladium/carbon at high pressure resulted in deprotection of the nitrogen and hydrogenation of the double bond. A spontaneous cyclization reaction closed the B ring and resulted in the formation of iminium ion 1.5.6. Reduction of the iminium ion gave 6-azaspirocycle 1.5.7 with the desired configuration at C5. A series of straightforward  10  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  transformations provided aldehyde 1.5.8. The C15-C21 portion was introduced by means of a Wittig olefination with phosphorane 1.5.9. The reaction proceeded slowly but in good yield to give dienone 1.5.10. Reduction under Luche conditions gave a 5:2 mixture of C17 epimers in favour of the desired 17R diastereomer. After this compound was converted into β-aminoaldehyde 1.5.12 a Horner-Wadsworth-Emmons olefination reaction with phosphonate 1.3.10 gave the desired tri-substituted alkene 1.5.13. This was subsequently used to make both pinnaic acid (II) and tauropinnaic acid (III). Scheme 1. 5 Heathcock’s Approach Towards Pinnaic Acid and Tauropinnaic Acid N C  EtO 2C  O  LiAlH4, Et2O 0 oC (94 %)  14  1.5.1  N  O  NHCBz  H  H  1.5.2  1.5.3  9  -50 oC to -20 oC (53 %)  13  HO  H 1.5.4 H  O  t-BuO 2C  t-BuO 2C H2, 55 psi Pd/C, EtOAc (87 %)  CbzHN AcO  H  N  t-BuO 2C  +  5  H  N  H  H  1.5.5  1.5.6 t-BuMe2SiO  H N  O  1.5.7 H  PPh3  1.5.9  t-BuMe2SiO  O  Cl  O  N  MeOH, 65 oC, 3 d (77%)  H  B  AcO  AcO  H  O  CbzHN  TiCl4, CH2Cl2  H  Cl O 1.5.8  1.5.10  H  O t-BuMe2SiO  NaBH4, CeCl3.7H2O, MeOH (90-92 %, 5:2 17R:17S)  O  N  HO 1.5.11  H Cl Et3SiO  H EtO 2C PO(OEt) 2 1.3.10 LiCl, DBU, MeCN (54 %)  EtO 2C HN OSiMe 2t-Bu  Pinnaic Acid and Tauropinnaic Acid H  Cl Et3SiO  1.5.13  HN  H  17  Cl  t-BuMe2SiO  H  1.5.12  11  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.4.2.6 Heathcock’s Completion of Halichlorine  2e  Scheme 1. 6 Heathcock’s Completion of Halichlorine PhS  H  H  O  MeO2C PO(OMe) 2 1.6.1  HN OSiMe 2t-Bu  PhSLi, THF 0 oC to rt (71 %)  H Cl Et3SiO  MeO2C HN OSiMe 2t-Bu  1.5.14  HN OSiMe 2t-Bu  + H  Cl Et3SiO  CO2Me H  PhS  H  Cl Et 3SiO  1.6.2a  DMF, 55 oC 35 h (48-61 %)  t-BuMe2SiO  MeO2C  Cl  H  H  PhS  K2CO3, PhSH  1.6.2b  A HN  t-BuMe2SiO  Cl OSiEt 3 1.6.3  H  HO  NaO2C  Cl  HO 1.6.5  N  MeO2C  N  OSiEt 3 1.6.4  EDCI.HCl DMAP DMAP.HCl CHCl3, THF Reflux (32 % over 2 steps)  H O  N O  D  Cl  HO halichorine (I)  Aldehyde 1.5.12 was also used as an intermediate in a synthesis halichlorine (I). The aldehyde was converted into a mixture of double bond isomers 1.6.2a and 1.6.2b by treating with trimethylphosphonoacrylate (1.6.1) and lithium phenylthiolate (Scheme 1. 6). Heating this mixture in the presence of benzenethiol and potassium carbonate closed the A ring to give tricycle 1.6.4. Presumably this transformation proceeds through intermediate 1.6.3, the result of conjugate addition of phenylthiolate to the α,β-unsaturated ester followed by elimination of the 1° thioether in favour of the 2° thioether. Subsequent intramolecular Michael addition of the amine followed by elimination of phenylthiolate gave the desired product. After removal of the protecting groups and saponification of the methyl ester, macrolactonization using Keck’s conditions (see Scheme 1.6 for details) gave halichlorine (I). 6 The Yamaguchi protocol for macrolactonization was attempted but resulted in a lower yield. This was attributed to the  12  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  presence of the free 2° hydroxyl group. The Heathcock group was able to obtain a crystal structure of (±)-halichlorine (I) and confirm the relative configuration reported by Uemura and by Danishefsky. 1.4.2.7 Uemura/Arimoto Approach Towards Pinnaic Acid 7 Scheme 1. 7 Uemura and Arimoto’s Approach Towards Pinnaic Acid EtO 2C t-BuO 2C  14  13  EtO 2C  CF3CO2H, CH2Cl2 (97 %)  C  (Me3Si)2NK, THF, -15 oC; Br, -15 oC  HO2C  (81 %)  H  H 1.7.1  EtO 2C 9  HO2C H  1.7.2  1.7.3 O  t-BuPh2SiO (PhO)2PON3, NEt3 CbzHN PhH, reflux; i BnOH, Pr2NEt, DMAP PMPO reflux H (88 %)  HO2C PMPO  H 1.7.4  PMPO H 1.7.6  1.7.5 H  t-BuPh2SiO  5  H  N  H  H O B F3COC  PMPO  N  PMPO H  EtO 2C PO(OEt) 2 1.3.10 NaH, THF (80 % over 2 steps)  EtO 2C F 3COC  N  PMPO H  H 1.7.8  1.7.7  20 % Pd(OH)2, H2 cat. AcOH, EtOH (93 %)  CbzHN  1.7.9  H EtO 2C F 3COC  N  Pinnaic Acid (II)  O H 1.7.10  Reminiscent of the approaches taken by both the Danishefsky and Heathcock groups, the Uemura approach to this family of natural products began with a highly substituted cyclopentane ring 1.7.1 (Scheme 1. 7). This known racemic diester 8 contains the C ring and has the desired relative configuration for C13 and C14. Selective hydrolysis of the tert-butyl ester followed by alkylation with allyl bromide established the C9 spirocenter. A series of efficient transformations led to carboxylic acid 1.7.4, which after treatment with standard Curtius rearrangement conditions was heated in the presence of benzyl alcohol to provide protected amine 1.7.5. Elaboration of the allyl side chain gave enone 1.7.6, which under hydrogenation conditions,  13  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  similar to those used by Heathcock, gave 6-azaspirocycle 1.7.7. The nitrogen was then protected as its trifluoroacetate derivative and the silylated 1° alcohol was converted into the corresponding aldehyde 1.7.8. A Horner-Wadsworth-Emmons olefination with phosphonate 1.3.10 gave α,β-unsaturated ester 1.7.9. This compound was converted into aldehyde 1.7.10, which is a racemic version of a compound used in the Danishefsky approach towards pinnaic acid (II). 1.4.2.8 Uemura’s Synthesis of the Tricyclic Core of Halichlorine7 Scheme 1. 8 Uemura and Arimoto’s Approach Towards Halichlorine H  MEMO  MEMO  CbzHN  H PMPO  1.8.2  N  Proton Sponge, 60 oC (60 %)  PMPO  H  H  Br  N PMPO H  H  1.7.5  1.8.1 12 mol %  Mes N  Grubbs II  H  1.8.3 N Mes Cl Ru  Cl  PCy3 ethylene, PhCH3, 80 oC (72 %)  N PMPO H 1.8.4  halichlorine (I)  H  Ph  N PMPO H 1.8.5  H O  N  PMPO H 1.8.6  Mes =  Starting with the pinnaic acid (II) precursor 1.7.5, the Uemura group was able to elaborate to tricycle 1.8.4, a structure which contains a significant portion of halichlorine (I) (Scheme 1. 8). The Uemura group synthesized 6-azaspirocycle 1.8.1 using chemistry parallel to that used to make 1.7.7 in their pinnaic acid synthesis. Alkylation with propargyl bromide (1.8.2) gave alkyne 1.8.3 which was subsequently used to make enyne 1.8.4. Closure of the A ring was accomplished by treatment with the second generation Grubbs catalyst (Grubbs II) under an ethylene atmosphere which promoted a smooth enyne metathesis reaction to give diene 1.8.5. It should be noted that the presence of the tertiary amine did not have a deleterious effect on this reaction; tertiary amines have been known to be incompatible with Grubbs-type  14  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids 9  catalysts. The terminal olefin of 1.8.5 was converted into an aldehyde to give compound 1.8.6. This compound represents the most advanced intermediate made by the Uemura group towards halichlorine. It should be noted that the Uemura group has investigated an alternative asymmetric route towards these natural products however the configuration at C14 turned out to be incorrect.10 1.4.2.9 Contributions from the Zhao and Ding Laboratories 10 Scheme 1. 9 Zhao and Ding’s Approach Towards Halichlorine and Pinnaic Acid O O  H 13  LDA, MeI, THF, -78 oC (77 %)  C  H 1.9.1  O O  H  14  BnO  H 1.9.2  O 2N BnO  CO2Me  H  urea, MeCN, 80 oC (76 %)  1.9.3  MeO2C H  1) H2NOH.HCl, K2CO3 MeOH, (97 %) 2) m-CPBA, Na2HPO4  O  t-BuMe2SiO  O 1) Ni2B, N2H4.H2O, EtOH, reflux (72 %) 2) NaBH4, MeOH  O 2N 9  Triton B, t-BuOH THF, (95 %)  H  BnO  O 2N  H BnO  1.9.4  1.9.5 H  t-BuMe2SiO  5  HO  N  HO  H  F3COC  H  N  F 3COC  BnO  pinnaic acid  N  HO H  H 1.9.7  H 1.9.6  EtO 2C  B  BnO  0 oC to rt (96 %)  1.9.8  H 1.9.9  halichlorine  Zhao, Ding and co-workers have published an approach towards halichlorine and pinnaic acid that starts with the bicyclic lactone 1.9.1 (Scheme 1. 9). This lactone contains the C ring of halichlorine and has the C13 stereocenter set. The C14 methyl group was installed by stereoselective alkylation with methyl iodide on the more accessible exo face of the bicycle. The resulting lactone was converted to ketone 1.9.3, which was then transformed into an oxime and oxidized with meta-chloroperoxybenzoic acid (m-CPBA) to give nitrocyclopentane 1.9.4. The nitro group would later become the requisite nitrogen. Deprotonation of nitrocyclopentane 1.9.4 and ensuing Michael addition with methyl acrylate provided ester 1.9.5 stereoselectively, which  15  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  after a series of transformations was converted into nitro ketone 1.9.6. Reduction of the nitro group with nickel boride resulted in ring closure with the pendant ketone forming an isolable nitrone intermediate that was then reduced with sodium borohydride to give hydroxyl amine 1.9.7. This sequence closed the B ring and established the C5 stereocenter. 6-Azaspirocycle 1.9.7 was used to make hydroxy-trifluoroacetamide compound 1.9.8. This compound closely resembles compound 1.1.7 (see Scheme 1. 1) used in the Danishefsky approach towards halichlorine and it also resembles a compound used in Kibayashi’s approach towards halichlorine. The Zhao and Ding group also synthesized α,β-unsaturated ester 1.9.9, a compound which intercepts an intermediate used in Danishefsky’s synthesis of pinnaic acid. 1.4.2.10 Contributions from the Martin Laboratory 11 Scheme 1. 10 Martin’s Approach Towards Pinnaic Acid MeO2C  OLi 1.10.2  N  THF, -78 oC  1.10.1  MeO2C O N  I  -  MeO2C  1.10.4 H  THF, DMPU -78 oC to rt (68 %)  (H8C4)NOC  t-BuPh2SiO  H 1.10.6  EtO 2C 1.10.7 10 mol % Grubbs II CH2Cl2 (dr 10:1)  C  H  1.10.3  BocHN  9  13 14  1.10.5  EtO 2C  pinnaic acid  BocHN t-BuPh2SiO  H  1.10.8  The Martin approach towards pinnaic acid started with the conversion of 1.10.1 to 1.10.5. This sequence began with a Michael addition between Z-enolate 1.10.1 and methylcyclopentenoate (1.10.2) (Scheme 1. 10). The intermediate anion 1.10.3 was then trapped with 5-iodo-1-pentene (1.10.4) to give the highly functionalized cyclopentane 1.10.5. This sequence introduced the C ring and established the desired configurations at C9, C13 and C14. This transformation is based upon work done previously in the Heathcock lab. 12,13 Following straightforward transformations cross-metathesis of alkene 1.10.6 with diene 1.10.7 in the presence of a second generation Grubbs catalyst gave diene 1.10.8. This compound is the racemic form of compound 1.2.3, a compound made in Danishefsky’s approach towards pinnaic acid (see Scheme 1. 2).  16  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.4.2.11 The Martin Approach Towards Halichlorine Scheme 1. 11 The Martin Approach Towards Halichlorine H OHC BocHN  BocHN t-BuPh2SiO  t-BuPh2SiO  H 1.10.6  ii) aqueous K2CO3, 0 oC (89 %) H  NMO, DIBAL-H; then 1.10.5 THF, (dr 2:1)  t-BuPh2SiO  B  H 1.11.2  H  EtO 2C  Ac2O, NEt3  OH HN t-BuPh2SiO  9  HN  1.11.1 H  CO2Et  OHC  i) CF3CO2H, CH2Cl2, rt  DMAP, CH2Cl2 (40 % 2 steps) H  1.11.3  A EtO 2C  N  t-BuPh2SiO  halichlorine  H 1.11.4  The Martin group also completed a formal total synthesis of halichlorine from alkene 1.10.6 (Scheme 1. 11). Elaboration to enal 1.11.1 was followed by conversion to 6-azaspirocycle 1.11.2 through a sequence involving deprotection of the Boc group with trifluoroacetic acid followed by intramolecular Michael addition under basic conditions. This sequence, which closed the B ring and set the C9 stereocenter, is similar to the conversion of 1.1.5 to 1.1.7 (Scheme 1. 1) in Danishefsky’s approach towards halichlorine. Hydroalumination of ethyl propiolate and subsequent addition to the aldehyde function of 1.11.2 gave allylic alcohols 1.11.3 as a 2:1 mixture of diastereomers. Treatment of this mixture with acetic anhydride, triethylamine and N,N-dimethylaminopyridine (DMAP) resulted in an intramolecular cyclization reaction that closed the A ring and provided tricycle 1.11.4. This compound is the same as compound 1.4.4 synthesized in Kibayashi’s approach towards halichlorine (see Scheme 1. 4). 1.4.2.12 Contributions from Pilli’s Laboratory 14 The Pilli approach towards halichlorine and the pinnaic acids was published just before the work published by the Martin group and begins with a very similar opening sequence as that reported by the Martin group (Scheme 1. 12). In this case, intermediate anion 1.10.3 was trapped with ethyl-4-iodobutanoate (1.12.1) to give diester 1.12.2. As in the Martin approach, this sequence established the C ring and introduced the desired configuration at C9, C13 and C14. Dieckmann cyclization followed by decarboxylation gave cyclopentanone 1.12.3. This  17  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  compound was modified slightly to give cyclopentanone 1.12.4 that was then treated with standard Beckmann rearrangement conditions to provide spirocyclic lactam 1.12.5 as the major product along with small amounts (~ 15 %) of undesired ring-opened products 1.12.6. The Beckmann rearrangement protocol expanded the 5-membered ring to a 6-membered ring and thus formed the B ring. 6-Azaspirocycle 1.12.5 is the most advanced intermediate reported by the Pilli group. Scheme 1. 12 Pilli’s Approach Towards Halichlorine and Pinnaic Acid MeO2C  MeO2C O  OLi 1.10.2  N  THF, -78 oC  N  1.10.1  EtO 2C MeO2C  I  1.12.1 THF, DMPU -78 oC to rt (68 %)  H  9  13 14  (H8C4)NOC  C  H  1.10.3  1.12.2  O  i) t-BuOK, PhCH3, reflux 5 min ii) H2SO4, THF, 4h (61%)  EtO 2C  -  O  (H8C4)NOC  t-BuPh2SiO  H  H  1.12.3  1.12.4 O  1) H2NOH.HCl, NaOAc, MeOH (88 %)  HN  2) TsCl, pyr., 0oC, 49 h t-BuPh2SiO  B  +  NC t-BuPh2SiO  H  1.12.5 (60 %)  1.12.6 (15 %)  1.4.2.13 Contributions from the Forsyth Laboratory 15 The Forsyth approach towards pinnaic acid started with the racemic aldehyde 1.13.1, which is available by oxidation of the corresponding alcohol (Scheme 1. 13). 16 This compound contains the C ring and has the desired configuration at C9 and C13; unfortunately the C14 configuration is incorrect. Treatment of this compound with acid resulted in formation of the desired iminium ion 1.13.2 which then reacted with allyltrimethylsilane to give 6-azaspirocycle 1.13.3. The stereochemical outcome is presumably the result of pseudo-axial attack of the nucleophile on iminium ion 1.13.2, where the largest substituent is in a pseudo-equatorial orientation in the half-chair conformation 1.13.2. 6-Azaspirocycle 1.13.3 was elaborated to the α,β-unsaturated ester 1.13.4. It should be noted that this molecule has the incorrect relative configuration at C5 and C14 required for pinnaic acid.  18  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  Scheme 1. 13 Forsyth’s Approach Towards Pinnaic Acid O  SiMe 3  BnOCOHN  +  N  CF3CO2H, CH2Cl2 -42 oC, 5 min (70 %)  C  AcO H  H  H  R BnOCO PG  MeO2C  B  5  BnOCO  AcO  AcO  N 14  H  Nuc  1.13.1  N  1.13.2  H  1.13.3  1.13.4  1.4.2.14 Contributions from the Wright Laboratory 17 Scheme 1. 14 The Wright Approach Towards Halichlorine O BnO  5  NH2  +  1.14.1 Grubbs II p-TsOH.H2O 56 h (83 %)  C  PhH reflux -H2O (96 %)  1.14.2 BnO  BnO  BnO N  1.14.3 HN  MgBr  HN  THF (76 %) 1.14.4  B  1.14.5  The Wright group has developed an approach towards the 6-azaspirocyclic ring system that incorporates a ring closing metathesis reaction to form the B ring (Scheme 1. 14). Chiral amine 1.14.1 is available from cysteine and has the desired C5 stereochemstry. Heating this amine with cyclopentanone (1.14.2) resulted in the formation of imine 1.14.3. Note that in this case cyclopentanone serves as a model for the B ring of halichlorine. Treatment of imine 1.27.3 with allylmagnesium bromide gave diene 1.14.4. A ring closing metathesis reaction with a second generation Grubbs catalyst formed the C ring. The reaction proceeded smoothly in the presence of p-toluenesulfonic acid to provide 6-azaspirocycle 1.14.5. Protonation of the nitrogen is necessary for the success of this reaction; otherwise the free amine will poison the catalyst resulting in poor yields. To date, this approach has only been tried on the model substrate shown in Scheme 1.14.  19  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.4.3  Groups That Start with the B Ring  1.4.3.1 Clive’s First Generation Approach Towards Halichlorine 18 Scheme 1. 15 Clive’s First Generation Approach Towards Halichlorine t-BuMe2SiO  MeO2C  5  (Boc)2N  O  CO2Me  1.15.1 from L-glutamic acid  O  O  Bn  OBn  MeO2C  O 1.15.2  O  1) BuLi, THF - 78 oC (78 %) 2) Dess-Martin periodinane CH2Cl2 (90 %)  OBn  13  1.15.4  t-BuMe2SiO O  SO 2Ar  1.15.3  CHO  N O  NH  t-BuMe2SiO H N H  O SO 2Ar OBn  MeO2C  (Ph3P)4Pd dimedone THF (88 %)  1.15.5  HN  B SO 2Ar OBn  MeO2C 1.15.6  t-BuMe2SiO H  t-BuMe2SiO H  t-BuMe2SiO H Bu3SnH, AIBN HN MeO2C  SO2Ar Br  1.15.7  PhCH3, 75 oC (57 %)  HN  HN MeO2C  SO2Ar C  H 1.15.8  MeO2C  H 1.15.9  The first generation approach from the Clive laboratories involved the coupling of chiral sulphone 1.15.3 with chiral aldehyde 1.15.4 (Scheme 1. 15). Chiral sulphone 1.15.3 is available from L-glutamic acid and contains the C5 stereocenter. The C13 stereocenter is established in aldehyde 1.15.4 by an alkylation reaction with the chiral oxazolidinone 1.15.2. The coupling reaction was accomplished by deprotonation of the sulphone and adding it to aldehyde 1.15.4. The resultant diastereomeric alcohols were oxidized with Dess-Martin periodinane to provide the corresponding ketone 1.15.5. Palladium-catalyzed removal of the allyl carbonate resulted in cyclization of the B ring and formation of enamine 1.15.6. Following conversion of the benzyl ether to bromide 1.15.7, exposure to radical conditions closed the C ring to give the 6-azaspirocyclic ring system 1.15.8. Cleavage of the sulfone gave 6-azaspirocycle 1.15.10 the most advanced intermediate in the Clive group’s first generation approach towards halichlorine and the pinnaic acids.  20  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.4.3.2 Clive’s Second Generation Approach Towards Halichlorine and the Pinnaic Acids18 Scheme 1. 16 Clive’s Second Generation Approach Towards Halichlorine Ph  Ph -  HN N Li H H Ph Ph 1.16.2 THF, Br  MeO2C Bn  N CO2Me  +  MeO2C  MOMO O  5  B  N Bn 9 MeO2C  (69 %)  1.16.1  AcO  1.16.3 MOMO O  Bu3SnH AIBN, PhH 80 oC  N  +  MOMO O  N Me2CuLi, Me3SiCl  13  1.16.5 (67 % from 1.16.4)  SePh 1.16.4  C  AcO  N  MOMO O  N  o  NEt3, THF, -78 C to rt (81 %)  1.16.6 (20 % from 1.16.4)  14  1.16.7  1) NaOMe, MeOH (91-92 %) 2) MeSO2Cl, NEt3, THF (64-67 %) 3) DBU, PhCH3, reflux (64-69 %) AcO MeO2C  H O  N  i) Me3OBF4, CH2Cl2 ii) 20 % aq Na2CO3 MeCN (72-77 %)  A MeO2C  N  MeO2C 1.16.8  1.16.9  The Clive group’s second-generation approach to halichlorine (I) began with the known meso diester 1.16.1, which contains the A ring (Scheme 1. 16). Alkylation with allylbromide in the presence of the chiral amide base 1.16.2 gave optically pure diester 1.16.3. This reaction established both the C5 and C9 stereocenters. Following elaboration to bicycle 1.16.4, heating in the presence of tri-n-butyltin hydride and azo-bisisobutyronitrile (AIBN) induced radical cyclization of the C ring to give 20 % of the desired enamide 1.16.6 and 67 % of α-acetoxy lactams 1.16.5. The mixture of α-acetoxy lactams 1.16.5 could be converted into enamide 1.14.6 through a simple 3-step procedure (refer to Scheme 1. 16 for details). The C14 methyl group required for halichlorine (I) was selectively installed by a conjugate addition reaction of  21  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  lithium dimethyl cuprate to give piperidinone 1.16.7. A straightforward sequence of reactions produced a mixture of allylic acetates 1.16.8. Opening of the lactam with Meerwien’s reagent (Me3OBF4) gave an intermediate amine which spontaneously cyclized and eliminated acetate ion when exposed to aqueous sodium carbonate to give tricycle 1.16.9. This compound represents a key advanced intermediate in the Clive group’s second generation approach towards halichlorine (I). 1.4.3.3 Clive’s Third Generation Approach Towards Halichlorine (1)18 The Clive group developed a third approach towards the halichlorine family starting with the reduced version of diester 1.16.1 (Scheme 1. 17). Diol 1.17.1 was easily converted into diene 1.17.2 that when warmed in the presence of a second generation Grubbs catalyst closed the C ring to form 6-azaspirocycle 1.17.3. After elaborating to the tricyclic α,β-unsaturated ester 1.17.4, reduction with lithium borohydride provided primary alcohol 1.17.5 as the sole product and set both the C13 and C14 stereocenters. The primary alcohol was converted to the primary iodide and subsequent radical cleavage gave lactam 1.17.6 with the requisite methyl group in place. This compound was used to prepare alcohol 1.17.7 that has been reported (in racemic form) in the Kibayashi approach towards halichlorine (I) and pinnaic acid (II). Scheme 1. 17 Clive’s Third Generation Approach Towards Halichlorine  Bn  t-BuMe2SiO  OSiMe 2t-Bu  OH N  Bn  HO  2 mol % Grubbs II CH2Cl2, 35 oC (85-95 %)  N  Bn  H  O  MeO2C 1.17.4 H t-BuMe2SiO N O H 1.17.6  1) I2, Ph3P, imidazole,  t-BuMe2SiO  LiBH4, THF MeOH (87 %)  N  C  1.17.3  H t-BuMe2SiO  N  HO  OH 1.17.2  1.17.1  H  O  CH2Cl2 (96 %)  N 14  13  2) Bu3SnH, AIBN, PhH reflux (99 %)  HO H 1.17.5 H  HO O  N  halichlorine (I) pinnaic acid (II) H  1.17.7  22  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.4.3.4 The Simpkins First Generation Approach Towards Halichlorine  19  The Simpkins first approach towards halichlorine and pinnaic acid started with the asymmetric allylation of the known diester 1.18.1 (Scheme 1. 18). The product of this reaction, piperdine 1.18.2, contains the B ring of halichlorine and has the desired configuration for C5 and C9. Elaboration of the product to diene 1.18.3 was followed by ring closing metathesis to form the C ring and gave allylic alcohol 1.18.4. Upon conversion to azaspirocyclopentanone 1.18.5 treatment with vinylmagnesium bromide in the presence of cerium trichloride provided allylic alcohol 1.18.6. This compound appears to have the incorrect configuration required for C13 therefore this route would have to be modified in order to make halichlorine.  Scheme 1. 18 The Simpkins First Approach Towards Halichlorine and Pinnaic Acid Ph  MeO2C  OSiMe 2t-Bu  Ph MeO2C  N Bn MeO2C 1.18.1 t-BuMe2SiO  Bn  H N  HO 1.18.4  NLi LiN Ph  o  THF, - 78 C Br  Ph  N Bn MeO2C  (77 %)  Bn  OH 1.18.3  1.18.2 t-BuMe2SiO  Bn  H N O  1.18.5  N  MgBr  t-BuMe2SiO  CeCl3, THF -78 oC (41 %)  Bn  2 mol % Grubbs II CH2Cl2, 20 oC (80 %)  H N  HO 1.18.6  1.4.3.5 The Simpkins Second Generation Approach Towards Halichlorine19 The Simpkins group also investigated a second approach towards halichlorine and pinnaic acid (Scheme 1. 19). Diester 1.18.2 was elaborated to allylic alcohol 1.19.1. Heating this compound with triethylorthoacetate in the presence of a catalytic amount of acid formed the intermediate allyl vinyl ether 1.19.2 that subsequently underwent a Johnson-Claisen rearrangement to give a 4:1 mixture of the desired diene 1.19.3 and its C13 epimer. Ring closing metathesis closed the C ring to give 6-azaspirocycle 1.19.4. This compound represents the Simpkins most advanced intermediate towards halichlorine and pinnaic acid.  23  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  Scheme 1. 19 The Simpkins Second Approach Towards Halichlorine and Pinnaic Acid MeO2C Bn  5  N  t-BuPh2SiO  B  Bn  9  OEt H  CH3C(OEt)3  N  CH3CH2CO2H (cat)  t-BuPh2SiO  120 oC  MeO2C  N  O  Bn  OH 1.18.2  1.19.1  t-BuPh2SiO Bn  Grubbs II, PhCH3  N  t-BuPh2SiO  o  80 C (46 % from 1.19.1)  N  Bn  13  H  1.19.2  H  C  EtO 2C  H  CO2Et  1.19.3 plus C13 epimer (ca 4:1)  1.19.4  1.4.3.6 Contributions from the White Group 20 Scheme 1. 20 White’s Approach Towards Halichlorine O  O HN3, NEt3 PhH, heat (85 %) O  O  O  O TsOH, aq. MeOH  N3  O O  O  1.20.1  O  N O  1.20.2  O  1.20.3  O O  +B  O HO  N H  O  H  O  5  N -  PhMe, reflux (64 % from E isomer)  O  O  N 9  H O  O  1.20.4  1.20.5  1.20.6  H MeO2C HN HO HO H 1.20.7  The approach from the White laboratories began with the synthesis of triene 1.20.1 (Scheme 1. 20). The nitrogen found in halichlorine was introduced by treatment of 1.20.1 with hydrazoic acid to give azide 1.20.2. After elaboration to the macrocyclic oxaziridine 1.20.3,  24  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  exposure to p-toluenesulfonic acid (TsOH) in aqueous methanol released oxime 1.20.4, which spontaneously cyclized to form the B ring and gave nitrone 1.20.5. Heating the nitrone in toluene resulted in a [3+2] cycloaddition reaction to close the C ring and yielded tetracycle 1.20.6. The White group subsequently elaborated this compound to amino diol 1.20.7, the most advanced intermediate published by White and co-workers. 1.4.3.7 Contributions from the Feldman Group 21 Scheme 1. 21 The Feldman Approach Towards Halichlorine 5  HN  Bu3Sn  B  O  PhI(CN)(OTf)  N  Bu3Sn O  N  9  SnBu3  SnBu3  SnBu3 1.21.1  IPh(OTf)  1.21.2  1.21.3 Bu3Sn  Bu3Sn N  O  TolSO2Na  (58-65 %)  TolO 2S  SnBu3  1.21.4  O  1.21.5  i) Li, naphthalene THF, -78 oC  N  ii) MeI, -60 oC (55 %)  C  MgBr2, PhCH3 reflux (69 %)  TolO 2S SnBu3  Bu3Sn  N  O  TolO 2S  Bu3Sn O  HN  N 14  13  H  1.21.6  H  1.21.7  t-BuPh2SiO  H 1.21.8  H  EtO 2C  N  Halichlorine (I)  HO H 1.21.9  The Feldman approach began with racemic bis-allylated dehydro-piperidine 1.21.1 (available from the bis allylation of pyridine) (Scheme 1. 21). It contains the B ring and has the C5 and C9 configurations established. After further elaboration to alkynylstannane 1.21.2, exposure to Stang’s reagent [PhI(CN)(OTf)] presumably formed iodonium salt 1.21.3.  25  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  Treatment with sodium p-toluenesulfonate in refluxing dimethoxyethane gave the product of carbene insertion, bicycle 1.21.5, via intermediate 1.21.4. In the presence of magnesium bromide, the β position of the ene-lactam was sufficiently activated to engage the proximal C-Sn bond to close the C ring and form tricycle 1.21.6. Reductive cleavage of the p-toluenesulfonate group, followed by addition of methyl iodide installed a methyl group alpha to the lactam carbonyl (1.21.6→1.21.7) with the desired C14 configuration. This compound was initially converted into 6-azaspirocycle 1.21.8, and then into tricyclic compound 1.21.9, a compound that intercepts a key intermediate in the Kibayashi approach towards halichlorine (I). 1.4.3.8 Contributions from the Ihara Laboratory 22 Scheme 1. 22 The Ihara Approach Towards Halichlorine and Pinnaic Acid O N  CO2t-Bu  B  H  Br  O Bu3SnH, AIBN PhH, reflux (78 %)  H  1.22.1  1.22.2  O  O  CO2t-Bu N  O N  Bn t-BuO 2C  H 1.22.4  1.22.3 H PtO2, H2 EtOH (100 %)  EtO 2C HN t-BuO 2C H 1.22.6  O  EtO 2C  LiNH2BH3  N  THF, 40 oC (59 %)  14  +  N  Bn t-BuO 2C  1:9  13  9  C  H 1.22.5 H  Et 3SiO  LDA, THF, -78 oC  5  HN  O  N  t-BuO 2C  Et3SiO  MeI, -78 oC (85 %)  H 1.22.8  H 1.22.7  H  Et 3SiO  CO2t-Bu N  H HN  HO  H  H  1.22.9  1.22.10  The Ihara approach towards halichlorine and pinnaic acid uses an unusual radical translocation/cyclization strategy to build the 6-azaspirocyclic core (Scheme 1. 22). Pipiridinone 1.22.1, which contains the B ring of halichlorine, was treated with tri-n-butyltin hydride and AIBN in refluxing benzene, which then generated an aryl radical by homolytic cleavage of the  26  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  bromine. Abstraction of a hydrogen radical from the 6-position of piperidinone 1.22.2 (5 atoms away) followed by 5-exo-trig cyclization onto the pendant alkene resulted in formation of 6-azaspirocycles 1.22.4 and 1.22.5 in a ratio of 9:1 favouring the desired spirocycle 1.22.5. Spirocycle 1.22.5 has the desired configuration for both C9 and C13. The lactam carbonyl was then converted into α,β-unsaturated ester 1.22.6 that after hydrogenation set the desired configuration at C5; a similar procedure for setting this configuration had been previously worked out by the Shishido group (see Scheme 1. 24 below 1.24.5→1.24.6).25 Amino diester 1.22.7 was elaborated to give lactam 1.22.8. Alkylation with methyl iodide installed the C14 methyl group with the desired configuration, a similar method was used by the Danishefsky group.2b,c,d Lactam 1.22.9 was then opened with lithium amide-borane complex to give 6-azaspirocycle 1.22.10, the most advanced compound in the Ihara synthetic approach towards halichlorine and pinnaic acid. 1.4.3.9 Contributions from the Laboratories of Mol and Bubnov 23 The laboratories of Mol and Bubnov have developed a model approach towards the spirocyclic ring system of halichlorine and the pinnaic acids that uses ring closing metathesis to form the C ring (Scheme 1. 23). Heating δ-valerolatam (1.23.1), as the B ring source, with triallylborane gave bis-allylated product 1.23.2. Prior to ring closing metathesis the nitrogen was protected with a variety of protecting groups (see Scheme 1. 23 for details) as the Grubbs catalysts are not compatible with free amines. Subsequent treatment of dienes 1.23.3 with the first generation Grubbs catalyst led to smooth intramolecular cyclization to close the C ring and form 6-aza-spirocycles 1.23.4. Scheme 1. 23 Mol and Bubnov’s Approach Towards Halichlorine and Pinnaic Acid ( N H 1.23.1  O  )3B  THF, reflux, 2h (79 %)  PG N H 1.23.2  N PG 1.23.3  1-5 mol % Grubbs I CH2Cl2, rt, 1-4 h (79-99 %)  N PG 1.23.4  PG = PhCH2, PhCO, CH3CO, CF3CO, t-BuCO  27  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.4.3.10 Contributions from the Keck Laboratory  24  An approach that uses radical cyclization to form the five-membered ring of halichlorine has been reported28 by Keck and Dalton at an ACS Meeting, but full details have not been published. 1.4.4  Groups that Make the B and C Rings in One Step  1.4.4.1 The Shishido/Itoh First Generation Approach Towards Halichlorine 25 Scheme 1. 24 The Shishido and Itoh First Generation Approach Towards Halichlorine MeO2C  MeO2C H  H2NOH.HCl, NaOAc EtOH, reflux (90 %)  O  1.24.1  MeO2C O  +  -O N  N  1.24.2  1.24.3 H  5  MeO2C O  4 Steps B  N  MeO2C  H2, 5 atm,  HN  9  PtO2, MeOH (100 %)  t-BuPh2SiO  C  H  H 1.24.4  1) KOH, EtOH 2) EDCl. HCl, HOBT NEt3, CH2Cl2 (61 %)  HN t-BuPh2SiO H 1.24.7  t-BuPh2SiO  1.24.6  H  H  EtO 2C  5  HN  H  1.24.5  CN  MeO2C  A NC O t-BuPh2SiO  N  H N  NC H  O t-BuPh2SiO  1.24.8  H  1.24.9  The Shishido and Itoh laboratories used a [3+2] dipolar cycloaddition reaction to make the 6-azaspirocyclic core of halichlorine, an approach that is thematically related to White’s approach seen above (Scheme 1. 24). Ketone 1.24.1 was heated with hydroxyl amine·hydrochloride to produce an intermediate oxime 1.24.2 that after formation of nitrone 1.24.3 underwent spontaneous [3+2] dipolarcycloaddition to give tricycle 1.24.4. This sequence formed both the B and C rings and set the C9 stereocenter. Unfortunately, the methylene-ester substituent of the B ring had the incorrect relative configuration required for C5 of the natural products. To correct this problem a five step sequence was employed which culminated in the  28  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  hydrogenation of α,β-unsaturated ester 1.24.5 to establish the desired C5 configuration. This compound was elaborated to give β-cyano ester 1.24.7. Hydrolysis of the ester followed by treatment with standard amide-forming conditions closed the A ring to give tricycle 1.24.8. This compound was converted into α,β-unsaturated cyano compound 1.24.9, which has the tricyclic core of halichlorine and represents a key advanced intermediate in the Shishido/Itoh approach towards halichlorine and the pinnaic acids. 1.4.4.2 The Shishido/Itoh Second Generation Approach Towards Halichlorine25 Scheme 1. 25 The Shishido and Itoh Second Generation Approach Towards Halichlorine EtO 2C B O  EtO 2C  Baker's Yeast water, 30 oC (63 %)  HO  1.25.1  1) KOH, water 2) (PhO)2PON3 PhH, reflux (65 % from 1.25.4a) (82 % from 1.25.4b)  I RO 1.25.3a (Z alkene, R=PMB) 1.25.3b (E alkene, R=Bn) LDA, THF, HMPA (59 % for 1.25.3a) (75 % for 1.25.3b)  EtO 2C HO 1.25.4a (Z) 1.25.4b (E)  1.25.2  RO  HO O  HN O  H 1.25.5a (Z) 1.25.5b (E)  (NH4)2Ce(NO2)6, MeCN-water for 1.25.5a (86 %) or Li, liq. NH3for 1.25.5b (77 %)  O  HN O  H 1.25.6a (Z) 1.25.6b (E) H  Cl MsCl, DMAP, CH2Cl2, (92 % from 1.25.6a) (94 % from 1.25.6b)  RO  O  HN O  H 1.25.7a (Z) 1.25.7b (E)  5  NaH or KH, Ph3P or Ph2PCH2CH2PPh2, THF Pd2(dba)3.CHCl3  H  O  N  5  +  O H 1.25.8  N  O  A  O  (~1:1)  H 1.25.9  The Shishido/Itoh groups have also published an alternative optically pure synthesis towards halichlorine that was distinct from their earlier approach. This synthesis began with cyclopentanone 1.25.1, which contains the C ring of halichlorine (Scheme 1. 25). The keto ester 1.25.1 was reduced with Bakers yeast to give the optically pure alcohol 1.25.2. Alkylation with iodides 1.25.3a or 1.25.3b occurred exclusively from the face opposite to the hydroxyl group of cyclopentanol 1.25.2 to give hydroxyl esters 1.25.4a and 1.25.4b and thus set the C9 configuration. Ester hydrolysis was followed by treatment with standard Curtius rearrangement conditions to introduce the nitrogen atom required in halichlorine. The intermediate isocyanate  29  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  (not shown) was trapped internally by the secondary alcohol five atoms away and formed the cyclic carbamates 1.25.5a and 1.25.5b. Deprotection of the allylic ethers was followed by conversion to the corresponding chlorides 1.25.7a and 1.25.7b. Treatment with palladium (0) gave approximately 1:1 mixtures of 1.25.8 and 1.25.9. Tricycle 1.25.9, which has the desired configuration for C5, is the most complex compound synthesized by the Shishido/Itoh groups in this route towards halichlorine (I). 1.4.4.3 Contributions from the Zhao Laboratories 26 Scheme 1. 26 Zhao’s Approach Towards Halichlorine H2NOH.HCl, NaOAc xylene-water, reflux (92 %)  MeO2C O  MeO2C H  1.26.1 MeO2C O  H 1.26.4  O  1.26.2  1) Zn, AcOH, water (94 %) 2) o-Cl2C6H4, reflux 24 h (84 %)  N  MeO2C  +  -O N  N  1.26.3  H  MeO2C HN HO  H  1.26.5  The approach from the Zhao group (Scheme 1. 26) is very similar to the Shishido/Itoh approach depicted in Scheme 1.25 as it relies upon a [3+2] dipolar cycloaddition to form the 6-azaspirocenter. Heating ketone 1.26.1 with hydroxylamine⋅hydrochloride forms the intermediate oxime 1.26.2. Michael addition followed by regioselective [3+2] cycloaddition gave tricycle 1.26.4. This approach has an advantage over the Shishido/Itoh first generation approach in that the C14 methyl group is also set in the reaction sequence. As in the Shishido/Itoh approach this reaction sequence formed the B and C rings but the C5 configuration was incorrect. Unlike the Shishido/Itoh approach in which five reactions were used to correct this problem, the Zhao group was able to solve this problem in one step. Hence, treatment with zinc in aqueous acetic acid inverted the C5 side chain configuration. The N-O bond was then cleaved to give hydroxy-amine 1.26.5, the most advanced intermediate published by the Zhao group.  30  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.4.4.4 Contributions from the Stockman Laboratory  27  The Stockman group has also developed a conceptually identical approach to that of both the Shishido/Itoh groups and the Zhao group (Scheme 1. 27). When ketone 1.27.1 was heated with hydroxylamine·hydrochloride the intermediate oxime 1.27.2 underwent a Michael addition and [3+2] dipolarcycloddition to give tricycle 1.27.3. To correct the C5 configuration, tricycle 1.27.4 was first converted into amino-diol 1.27.5 and then heated with ethanol in a sealed tube to give the desired 6-azaspirocycle 1.27.6. This was further elaborated to the fully protected 6-azaspirocycle 1.27.7 and thus far is the Stockman group’s most advanced intermediate towards halichlorine. Scheme 1. 27 The Stockman Approach Towards Halichlorine EtO 2C O  EtO 2C  1.27.1 H EtO 2C  H EtO 2C  O EtO 2C  N  HN HO H  1.27.4  1.4.5  EtO 2C H  H2NOH. HCl, MeOH, (62 %)  EtO 2C  EtO 2C O  EtO 2C  1.27.2  EtOH, 120 oC (100 %)  +  -O N  N  1.27.3  H EtO 2C  H HO 1.27.5  H EtO 2C  HN  F 3COC O  HO HO  H  1.27.6  N  O  H  1.27.7  Groups That Synthesize the C15-C21 Side Chain  1.4.5.1 Contributions from the Weinreb Laboratory 28 The Weinreb group has developed a method to install the C15-C21 side chain by means of a Horner-Wadsworth-Emmons reaction with an appropriate aldehyde (Scheme 1. 28). Heating alkyne ester 1.28.1 with hydrochloric acid in the presence of copper (I) chloride gave Z alkene 1.28.2. Through a series of steps, the hydroxy-acid 1.28.3 was transformed into Weinreb amide 1.28.4. Treatment with Grignard reagent 1.28.5 gave phosphonate 1.28.6. A Horner-Wadsworth-Emmons reaction with an aldehyde should give the requisite E olefin and subsequent reduction of the ketone should provide the C17 alcohol.  31  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  Scheme 1. 28 Weinreb’s Synthesis of the C15-C21 Subunit of Halichlorine and Pinnaic Acid O OH  BrMgH2C P(OMe)2 THF, -78 oC N(Me)OMe (83%)  CO2H  CO2H  Cl  Cl  1.28.1  1.28.2  OSiMe 2t-Bu  Cl  OSiMe 2t-Bu  OH  CuCl, conc. HCl (60 %)  O  P(OMe)2 O  1.28.3 OSiMe 2t-Bu  RCHO, DBU LiCl, THF  R Cl  1.28.4  O  O  1.28.5  Reduce Ketone  OSiMe 2t-Bu 17  Cl  R  OH 1.28.6  1.4.5.2 Studies from the Taber Laboratories 29 Scheme 1. 29 Contributions from the Taber Group O Cl 1.29.1  RMgX, CuBr.SMe2 THF, -20 oC, (72-89 %) X = Br, Cl  R Cl OH 1.29.2  The Taber group has developed a method to synthesize Z-3-chloroallylic alcohols, a subunit of halichlorine and the pinnaic acids (Scheme 1. 29). Copper mediated addition of alkyl or aryl Grignard reagents to allylic-chloro-epoxides 1.29.1 provided Z-chloroallylic alcohols in good overall yields (~ 71-89 %). The Z:E ratios were typically higher than 10:1. These compounds could potentially be used to build the C15-C21 subunit of halichlorine and the pinnaic acids. 1.5  Conclusion Halichlorine and the pinnaic acids have attracted significant attention from the synthetic  organic community. At least twenty-one groups have published work towards this intriguing family of molecules. These molecules have proven to be challenging synthetic targets as several groups have had to investigate more than one synthetic route and a number of approaches remain unfinished. A number of total syntheses have been reported and several formal total syntheses have been done. The tricyclic core has garnered the most attention while a couple of groups have invested some time on the C14-C21 side chain. Of the approaches made towards the tricyclic core some groups have started with the C ring, others have started with the B ring while  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  32  others have tried to build both the B and C rings in one step. Regardless of the approach taken, the synthetic problems associated with these compounds have resulted in the achievement of exquisite total syntheses and the development of new reactions.  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids  1.6 1  33  References  a) Kuramoto, M.; Tong, C.; Yamata, K.; Chiba, T.; Hayashi, Y.; Uemura, D. Tetrahedron Lett.  1996, 37, 3867-3870. b) Chou, T.; Kuramoto, M.; Otani, Y.; Shikano, M.; Yazawa, K.; Uemura, D. Tetrahedron Lett. 1996, 37, 3871-3874. 2  a) Arimoto, H.; Kuramoto, M.; Hayakawa, I; Uemura, D. Tetrahedron Lett. 1998, 39, 861-862.  b) Trauner, D.; Schwarz, J. B.; Danishefsky, S. J. Angew. Chem. Int. Ed.1999, 38, 3542-3545. c) Carson, M. W.; Kim, G.; Hentmann, M. F.; Trauner, D.; Danishefsky, S. J. Angew. Chem. Int. Ed.2001, 40, 4450-4452. d) Carson, M. W.; Kim, G.; Danishefsky, S. J. Angew. Chem. Int. Ed.2001, 40, 4453-4456. e) Christie, H. S.; Heathcock, C. H. Proc. Nat. Acad. Sci. 2004, 101, 12079-12084. f) Hayakawa, I; Arimoto, H.; Uemura, D. Heterocycles. 2003, 59, 441-443. 3  Clive, D. L. J.; Yu, M.; Wang, J.; Yeh, V. S. C.; Kang, S. Chem. Rev.2005, 105, 4483-4514.  4  For a review on the synthesis of 6-azaspirocycles see: Dake, G. R. Tetrahedron 2006, 62, 3467-  3492. 5  a) Matsumura, Y.; Aoyagi, S.; Kibayashi, C. Org. Lett. 2003, 5, 3249-3252. b) Matsumura, Y.;  Aoyagi, S.; Kibayashi, C. Org. Lett. 2004, 6, 965-968. 6  Boden, E. P.; Keck, G. E. J. Org. Chem.1985, 50, 2394–2395.  7  a) Arimoto, H.; Asano, S.; Uemura, D. Tetrahedron Lett. 1999, 40, 3583-3586. b) Hayakawa, I.;  Arimoto, H.; Uemura, D. J. Chem. Soc, Chem. Commun. 2004, 1222-1223. 8  Yamaguchi, M.; Tsukamoto, M.; Hirao, I. Tetrahedron Lett. 1985, 26, 1723-1726.  9  a) Fu, G. C.; Nguyen, S. B.; Grubbs, R. H. J. Am. Chem. Soc. 1993, 115, 9856-9857. b)  Kirkland, T. A.; Lynn, D. M.; Grubbs, R. H. J. Org. Chem. 1998, 63, 9904-9909. 10  Zhang, H.-L.; Zhao, G.; Ding, Y.; Wu, B. J. Org. Chem. 2005, 70, 4954-4961.  11  Andrade, R. B.; Martin, S. F. Org. Lett. 2005, 7, 5733-5735.  12  Wallace, G. A.; Heathcock, C. H. J. Org. Chem. 2001, 66, 450-454.  13  Oare, D. A.; Henderson, M. A.; Sanner, M. A.; Heathcock, C. H. J. Org. Chem. 1990, 55, 132-  157. 14  de Sousa, A. L.; Pilli, R. A. Org. Lett. 2005, 7, 1617-1619.  15  Koviach, J. L.; Forsyth, C. J. Tetrahedron Lett. 1999, 40, 8529-8532.  16  Details of the preparation have been reported in a Ph.D. Thesis: Koviach, J. L. Thesis,  University of Minnesota, Minneapolis, MN, 1999. 17  Wright, D. L.; Schulte II, J. P.; Page, M. A. Org. Lett. 2000, 2, 1847-1850.  Chapter 1 Previous Work Towards Halichlorine and the Pinnaic Acids 18  34  a) Clive, D. L. J.; Yu, M.; Li, Z. Chem. Commun. 2005, 906-908. b) Clive, D. L. J.; Yeh, V. S.  C. Tetrahedron Lett. 1999, 40, 8503-8507. c) Yu, M.; Clive, D. L. J.; Yeh, V. S. C.; Kang, S.; Wang, J. Tetrahedron Lett. 2004, 45, 2879-2881. d) Clive, D. L. J.; Wang, J.; Yu, M. Tetrahedron Lett. 2005, 46, 2853-2855. 19  Huxford, T.; Simpkins, N. S. Synlett 2004, 2295-2298.  20  White, J. D.; Blakemore, P. R.; Korf, E. A.; Yokochi, A. F. T. Org. Lett. 2001, 3, 413-415.  21  Feldman, K. S.; Perkins, A. L.; Masters, K. M. J. Org. Chem. 2004, 69, 7928-7932.  22  Takasu, K.; Ohsato, H.; Ihara, M. Org. Lett. 2003, 5, 3017-3020.  23  Nieczypor, P.; Mol, J. C.; Bespalova, N. B.; Bubnov, Y. N. Eur. J. Org. Chem. 2004, 812-819.  24  Keck, G.; Dalton, S. A. Abstracts of Papers; 226th ACS National Meeting, New York, Sept 7-  11, 2003; American Chemical Society: Washington, DC, 2003, ORGN-187. 25  a) Shindo, M.; Fukuda, Y.; Shishido, K. Tetrahedron Lett. 2000, 41, 929-932. b) Yokota, W.;  Shindo, M.; Shishido, K. Heterocycles 2001, 54, 871-885. c) Itoh, M.; Kuwahara, J.; Itoh, K.; Fukuda, Y.; Kohya, M.; Shindo, M.; Shishido, K. Bioorg. Med. Chem. Lett. 2002, 12, 2069-2072. 26  a) Lee, S.; Zhao, Z. (S.) Org. Lett. 1999, 1, 681-683. b) Lee, S.; Zhao, Z. (S.) Tetrahedron Lett.  1999, 40, 7921-7924. 27  Arini, L. G.; Szeto, P.; Hughes, D. L.; Stockman, R. A. Tetrahedron Lett. 2004, 45, 8371-8374.  28  Keen, S. P.; Weinreb, S. M. J. Org. Chem. 1998, 63, 6739-6741.  29  Taber, D. F.; Mitten, J. V. J. Org. Chem. 2002, 67, 3847-3851.  Chapter 2 A First Generation Approach Towards Halichlorine…  35  2 Chapter 2 A First Generation Approach Towards Halichlorine and the Development of New Methods to form 6-Azaspirocycles1  1  Versions of this chapter have been published. See a) Hurley, P. B.; Dake G. R. “N-Bromosuccinimide Promoted Ring Expansion Reactions: Diastereoselective Formation of Functionalized Azaspirocyclic Cyclopentanones” Synlett 2003, 14, 2131-2134. b) Dake, G. R.; Fenster, M. D. B.; Hurley, P. B.; Patrick, B. O. “Synthesis of Functionalized 1-Azaspirocyclic Cyclopentanones Using Bronsted Acid or N-Bromosuccinimide Promoted Ring Expansions” J. Org. Chem. 2004, 69, 5668-5675.  Chapter 2 A First Generation Approach Towards Halichlorine…  2.1  36  Introduction Halichlorine and the pinnaic acids are synthetically challenging natural products that  have received much attention from the synthetic organic community. 1 To date, twenty-one groups have published work towards halichlorine. Of these, two total syntheses and six formal total syntheses have been reported. The C15 to C21 segment (see Figure 2. 1 for atom numbering) that makes up a significant portion of the D ring has not attracted much attention as this portion is arguably synthetically less challenging than the ABC tricyclic ring system. In fact, the groups that have completed the synthesis (either total or formal) use similar methods to introduce this segment and use a similar macrolactonization approach to close the D ring.1d,g In contrast, there has been a variety of methods developed towards the construction of the ABC tricyclic core of halichlorine. We decided to develop a unique approach to halichlorine that would incorporate a semipinacol reaction as the key step in making the 6-azaspirocyclic ring system. The plan for this approach will be given first, followed by the work that was done towards the synthesis of halichlorine. Throughout this chapter there will be instances where a specific reaction or concept requires some background information. The reader should expect to move back and forth between work that was done by me and the background information. Figure 2. 1 Numbering Scheme for Halichlorine 4  H  3  O 21  A 1 2  O  D  5  N H  B8 9  10  C  14 13 12  11  17  Cl  2.2  HO  Retrosynthetic Analysis Before attempting the total synthesis of halichlorine a plan was devised (Scheme 2. 1).  Our plan began with a simplification of the target structure to tricycle 2.1.1. As with other approaches towards halichlorine, the C14 to C21 side chain would be installed in the latter stages of the synthesis and the D ring would be formed last using the established macrolactonization protocols.1d,g Tricycle 2.1.1 represented a key synthetic intermediate in our strategy towards halichlorine and two possible synthetic routes towards this molecule were envisioned (Path A and Path B).  37  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 1 Retrosynthetic Analysis for Halichlorine H  A  O O  H  B  N  D H  A  C  RO2C  N  B C  O 2.1.1  Cl  HO halichlorine  Path A  Path B  H  H  N  R  N  RO2C OH  O  2.1.2  2.1.8  H N  R  O  EtO  +  Ts  SnMe3  H N  O 2.1.5a (R=H) 2.1.5b (R=CO2Et)  EtO  N  Ts  H R  N  O 2.1.6a (R=H) 2.1.6b (R=CO2Et)  H  N O 2.1.7  N  2.1.10  EtO  O  H  OH  O 2.1.9  2.1.3a (R=H) 2.1.4 2.1.3b (R=CO2Et)  R  H  Ts  H N  O  +  SnMe3 2.1.11  2.1.4  Path A would involve formation of the C ring last and introduction of the crucial spirocenter by means of a semipinacol rearrangement reaction. If successful this reaction would serve as a signature step in our approach towards halichlorine. The ring expansion substrate 2.1.2 could be made from the coupling of alkenyl stannane 2.1.3 and cyclobutanone (2.1.4). Alkenyl stannane 2.1.3 might be made from bicyclic lactam 2.1.5, which is the product of ring closing metathesis of diene 2.1.6. Diene 2.1.6 should be accessible from glutarimide (2.1.7). Path B involves formation of the A ring last by means of a ring closing metathesis reaction of diene 2.1.8, a compound that might come from the elaboration of 6-azaspirocycle 2.1.9. As with pathway A the spirocenter would be introduced by means of a semipinacol ring expansion reaction of the allylic cyclobutanol 2.1.10. The ring expansion substrate 2.1.4 might be made by a coupling reaction involving the known alkenyl stannane 2.1.11 and cyclobutanone  38  Chapter 2 A First Generation Approach Towards Halichlorine… 2  2  (2.1.7). Alkenyl stannane 2.1.11 can be made from glutarimide (2.1.7) in 4 steps. It should be noted that both pathways A and B begin with the same starting material, glutarimide. 2.3  Previous Work in the Dake Lab Before describing the details of our approach towards halichlorine it would be  informative to provide some background information regarding previous work carried out in the Dake lab. Specifically, work involving the formation of 6-azaspirocycles via semipinacol rearrangement reactions will be discussed. A semipinacol rearrangement is closely related to the pinacol rearrangement (Scheme 2. 2). The pinacol rearrangement derives its name from the starting material in the first reported example. 3 This reaction is the acid catalyzed conversion of pinacol (2.2.1) to pinacolone (2.2.2). Protonation of one of the alcohols of pinacol followed by dehydration forms a 3˚ carbocation intermediate 2.2.3. Migration of one of the methyl groups with concomitant carbon-oxygen double bond formation gives a new intermediate 2.2.4 which after loss of a proton provides pinacolone (2.2.5). The pinacol rearrangement may be defined in broader terms as the acid catalyzed rearrangement of a 1,2 diol to form an aldehyde or ketone. The reaction involves dehydration, migration of an alkyl, aryl or hydrogen atom and carbonyl formation. Scheme 2. 2 The Pinacol Rearrangement HO OH  2.2.1  2.2.5 +  H +  O  H2SO4  H2O OH  - H+ - H2O  2.2.2  H O  2.2.3  +  O H  2.2.4  The phrase semipinacol rearrangement originally referred to an unusual pinacol rearrangement in which 1,2 diol 2.2.6 was converted into ketone 2.2.8 (Equation 2. 1). 4 This example is unusual because the 1,2-diol ionized to form the secondary carbocation 2.2.7 rather than the expected tertiary carbocation.  39  Chapter 2 A First Generation Approach Towards Halichlorine…  Equation 2. 1 The Original Semipinacol Rearrangement R1 OH HO R2 R3  H+ - H2O  R1 HO  O +  R2 R3  2.2.6  2.2.7  R1  R3  R2 2.2.8  The term semipinacol rearrangement has since been changed and now refers to a larger class of rearrangement reactions that are reminiscent of the pinacol rearrangement. There are a number of requirements for a reaction to be labeled a semipinacol rearrangement. The first requirement is that the reaction involves the migration of an R group (R = alkyl, aryl or H) from an alcohol-bearing carbon to an adjacent carbon. The second requirement is that the adjacent carbon must either: a) contain a good leaving group, b) contain a latent leaving group or c) provide a pathway to form a carbocation. Finally, the alcohol-bearing carbon must form a carbonyl functional group. A semipinacol rearrangement is different from a pinacol rearrangement in that loss of water is not required as part of the rearrangement process. Another way of saying the same thing is that the starting material does not have to be a 1,2-diol. An example of a semipinacol rearrangement is the Tiffeneau-Demjanov rearrangement (Equation 2. 2) in which a β-amino alcohol is converted into a ketone. In this reaction an amine such as 2.2.9 is first converted into diazonium salt 2.2.10 that contains a good leaving group (N2). Migration of one of the alkyl substituents and carbon-oxygen double bond formation result in an overall ring enlargement process.  Equation 2. 2 A Tiffeneau-Demjanov Rearrangement O HO  2.2.9  NH2  HONO  HO  N2  2.2.10  +  2.2.11  Due to its vague definition, the semipinacol rearrangement encompasses a large number of reactions.4 For a more detailed review of this class of reactions the reader is directed to the doctoral thesis of Michaël Fenster and the references therein. 5 One of the projects studied in the Dake group is the total synthesis of natural products  40  Chapter 2 A First Generation Approach Towards Halichlorine…  that contain azaspirocyclic ring systems. Figure 2. 2 includes examples of natural products that contain an azaspirocyclic ring system (highlighted in blue). Figure 2. 2 Representative Alkaloids Containing 6-Azaspirocenters SCN O H13C6  HO  N  N  N  C6H13  O  H  H HO  fasicularin  OMe cephalotaxine  lepadiformine  One possible way to construct the azaspirocenter common to these natural products is through an acid catalyzed semipinacol rearrangement such as that shown in Scheme 2. 3.  Scheme 2. 3 Acid Catalyzed Semipinacol Rearrangement to Form 6-Azaspirocenters R  R  OH N ( )n Ts 2.3.1 n = 1,2,3...  + OH  N Ts  H+  ( )n  O R  R  OH N+ Ts  N Ts 2.3.2  ( )n  ( )n  Michaël Fenster, a former graduate student in the Dake group, found that acid could be used to promote the semipinacol rearrangement of allylic cyclobutanols 2.3.3 to 6-azaspirocycles 2.3.4 and 2.3.5 (Table 2. 1). 6 While this reaction is quite useful in certain cases, it does have some drawbacks. These reactions proceeded in modest to excellent yield and with poor to excellent diastereoselectivity. In most cases the diastereomeric products were inseparable from one another by chromatography. This made the stereochemical assignments of the products difficult. 7,8 The reactions were sluggish in some cases (entries 4, 5 and 6). Cyclobutanol 2.3.3a which bears a benzyl ether substituent decomposed under the reaction conditions. As well, the attempted ring expansion reactions of rings larger than 4-membered failed under these conditions.  41  Chapter 2 A First Generation Approach Towards Halichlorine…  Table 2. 1 Acid Catalyzed Semipinacol Rearrangement of Allylic Cyclobutanols R1  R1  R1  R2 acid T (oC), time (h)  N Ts  R2  CH2Cl2  OH  +  N Ts  2.3.3  R2 N Ts  O 2.3.4  O 2.3.5  Entry  Substrate  R1  R2  Acida  T (°C)  t (h)  Products  % yieldb  Ratiod  1  2.3.3a  OBn  H  CSA  45  13  decomposed  N/A  N/A  2  2.3.3b  OTBS  H  CSA  45  13  2.3.4b:2.3.5b  81  2.7:1  3  2.3.3c  OPNB  H  CSA  45  13  2.3.4c:2.3.5c  51c  1:1.8  4  2.3.3d  H  Ph  CSA  45  144  2.3.4d:2.3.5d  89  1:4.5  5  2.3.3d  H  Ph  HCl  0  48  2.3.4d:2.3.5d  93  1:14  6  2.3.3.e  H  Me  HCl  0  67  2.3.4e:2.3.5e  68  1:3.7  a  CSA = (+)-camphorsulfonic acid (1.2 equiv) in dichloromethane; HCl = hydrochloric acid (1.2 equiv) in dichloromethane. b Isolated yield. c Unreacted 2.3.3c was recovered in 32 % yield. d Ratios were determined by 1H NMR integration and/or GC analysis of the product mixture. The stereochemical outcome from the acid catalyzed semipinacol rearrangement of 2.3.3 can be rationalized through transition state analysis; substrate 2.3.3b will be used as an example in the following explanation (Scheme 2. 4). 9 The stereoselectivity is thought to derive from a transition state in which the substituent (OTBS) is pseudoequatorial. It is believed that there are two possible diastereomeric transition states: one in which the cyclic iminium ion intermediate reacts through a “half-chair” conformation (A) and one in which the cyclic iminium ion intermediate reacts through a “twist-boat” conformation (B). The “half-chair” conformation is lower in energy than the “twist-boat” conformation and results in the formation of the major diastereomer 2.3.4b.  42  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 4 Proposed Mechanistic Rationale for the Observed Diastereoselectivity in the Acid Mediated Semipinacol Rearrangement of 2.3.3b OTBS  Ts  N OH 2.3.3b  H+  +  OTBS  N  Ts  OH  OTBS  N  HO  +  Ts  + +  OTBS  N  OTBS  +  + +  Ts  HO  A  B  O  OTBS  N  Ts  N O 2.3.4b  OTBS  N Ts  O  OTBS  OTBS  Ts  Ts  +  N  HO  H+  Ts  N O 2.3.5b  It was hoped that the acid catalyzed semipinacol rearrangement reaction could be applied to the synthesis of a natural product. Specifically, Dr. Fenster was interested in synthesizing fasicularin. However, in order to synthesize fasicularin using a semipinacol-type ring expansion reaction, it was necessary to expand a five-membered ring to a six-membered ring. Because this transformation was not possible using acidic conditions, an alternative semipinacol reaction was developed. Hence, treatment of siloxy-epoxide 2.5.1 with titanium (IV) chloride gave the desired 6-azaspirocycle 2.5.2 as a single diastereomer (Scheme 2. 5). 10 This compound was used as a key intermediate in a formal synthesis of fasicularin. 11  43  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 5 The Siloxy-epoxide Semipinacol Rearrangement Approach Towards Fasicularin OTBS  Ts  OTBS H  TiCl4  N O OTMS  CH2Cl2 o  Ts  N  O Ts  2.5.2  SCN  OTBS  HO O  -78 C, 0.5 h (95 %)  2.5.1  OH  N  H13C6  N  fasicularin  When the halichlorine project was initiated, there were two semipinacol rearrangement methods available to synthesize the 6-azaspirocyclic core: the acid catalyzed method and the siloxy-epoxide method. The Bronsted acid method was attractive because it used relatively cheap reagents to give a high level of structural complexity. However, the Bronsted acid method was slow and often resulted in the formation of diastereomeric mixtures of products. In addition, the products were difficult to separate from one another and difficult to characterize. The siloxy-epoxide method was attractive because it exhibited high diastereoselectivity. Unfortunately this method would introduce at least two extra steps in the synthetic sequence (formation of the epoxide and protection of the 3˚ alcohol) and it would introduce another functional group (the hydroxyl group adjacent to the spirocenter) which would have to be removed at a later stage in the synthesis. Initially a synthesis of halichlorine would be attempted that would incorporate the acid catalyzed semipinacol rearrangement. If for some reason this method failed the strategy could be modified to incorporate the siloxy-epoxide semipinacol rearrangement. In order to attempt a semipinacol rearrangement reaction an appropriate substrate would have to be made. Initially Path A was explored. This would involve formation of the A ring by means of a ring closing metathesis reaction prior to attempting the semipinacol rearrangement reaction. 2.4  Formation of the A Ring of Halichlorine by Ring Closing Metathesis: Substrate Synthesis  2.4.1  Introduction Earlier in this chapter the retrosynthetic plan was described in which there were two  possible pathways to synthesize the tricyclic core of halichlorine (See Scheme 2. 1 above). In both pathways a semipinacol rearrangement would be used to establish the spirocenter. As well,  44  Chapter 2 A First Generation Approach Towards Halichlorine…  both pathways would form the A ring by means of a ring closing metathesis reaction (See Figure 2. 3 Below). In Path A this ring could be formed by a ring closing metathesis reaction of diene 2.1.6a or diene 2.1.6b while in Path B this ring could be formed by a ring closing metathesis reaction of diene 2.1.8. Path B differs from Path A in that the semipinacol rearrangement would be attempted prior to ring closing metathesis. Initial efforts were focused on Path A. Figure 2. 3 Formation of the A Ring of Halichlorine by a Ring Closing Metathesis Reaction Path A  Path B H  H N  R  RCM R  O 2.1.5a (R=H) 2.1.5b (R=CO2Et)  2.4.2  H  H  RCM N  RO2C O  2.1.6a (R=H) 2.1.6b (R=CO2Et)  N O 2.1.1  RO2C  N O 2.1.8  Path A: Synthesis of Dienes 2.1.6a and 2.1.6b Diene 2.1.6a could be synthesized in three steps from glutarimide (2.1.7) (Scheme 2. 6).2a  Glutarimide (2.1.7) was treated with sodium borohydride in ethanol to give the monoreduced intermediate (not shown). The pH of the reaction mixture was made acidic (pH~3) by the addition of hydrochloric acid in ethanol which resulted in the formation of 6-ethoxypiperidinone 2.6.1.2a Treatment of this compound with a Lewis acid (BF3·OEt2) produced an intermediate iminium ion which underwent nucleophilic attack by allyltrimethylsilane to give 6-allyl-piperidin-2-one (2.6.2). The lactam nitrogen was alkylated by treating with sodium hydride and allyl bromide to provide diene 2.1.6a. Alternatively, alkylation of the nitrogen could be done with ethyl-2-bromomethylacrylate to give diene 2.1.6b. Before describing the outcome of the ring closing metathesis reactions it would be useful to discuss the ring closing metathesis reaction in some detail.  45  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 6 Synthesis of Dienes 2.1.6a and 2.1.6b i) NaBH4  O H  EtO  o  EtOH, 0 C ii) HCl/EtOH pH=3 (72 %)  N O 2.1.7  H  H  H  TMS BF3 .OEt2, CH2Cl2  N  0 oC to rt (100 %)  O  H Br  H  NaH, THF, 0 oC (97 %)  N O  2.6.1  2.6.2  N O 2.1.6a  NaH, EtO 2C THF, 0 oC to rt (59 %)  Br  H  EtO 2C  N O 2.1.6b  2.5  Metathesis  2.5.1  Introduction In recent years, the metathesis reaction has become an important reaction in synthetic  organic chemistry. Equation 2. 3 shows an example of a ring closing metathesis reaction whereby the first generation Grubbs catalyst 2.6.4 is used to convert 1,6-heptadiene (2.6.3) into cyclopentene (2.6.5). Various research groups have devoted entire research programs to the development of new metathesis catalysts and ligands. 12 As well, this reaction has been reported in the total synthesis of numerous natural products. 13 The importance of this reaction was recognized by the awarding of the most recent Nobel Prize in chemistry to Professors Robert Grubbs, Richard Schrock and Yves Chauvin for their development of the metathesis reaction in organic synthesis.  Equation 2. 3 Example of Ring Closing Metathesis PCy3 Cl Ru Cl  Ph PCy3  +  2.6.4 2.6.3  2.6.5  2.6.6  46  Chapter 2 A First Generation Approach Towards Halichlorine…  2.5.2  Types of Metathesis Reactions Figure 2. 4 shows a variety of applications of the metathesis reaction. The examples  include: ring opening metathesis (ROM), ring-closing metathesis polymerization (ROMP), acyclic-diene metathesis polymerization (ADMET), cross metathesis (CM) and ring-closing metathesis (RCM). Figure 2. 4 Types of Metathesis Reactions 14 Ring Opening Metathesis (ROM)14a  PCy3 Cl Ru  O  Ph  Cl  OTBS OTBS  +  CN  TBSO  PCy3  OTBS CN  2.6.4  O 14b  Ring Opening Metathesis Polymerization (ROMP)  N H 3C  CF3  F 3C  O  F 3C Ph Me  H3 C  Mo  O Ph CF3  Ph Me n  2.6.7 Acyclic Diene Metathesis (ADMET)14c  2.6.7  N  N  n  Cross Metathesis (CM)14d N  N Cl Ru Ph  Cl BzO  ( )7  +  CO2Me  PCy3 2.6.8  Ring Closing Metathesis (RCM)14e  Ts N  2.6.4  Ts N  BzO  ( )7  CO2Me  47  Chapter 2 A First Generation Approach Towards Halichlorine…  2.5.3  Important Advances in Metathesis Catalyst Design Up until the 1980’s all metathesis reactions were done with poorly defined  multicomponent homogeneous and heterogenous systems. The first significant breakthrough in catalyst design was the development of the molybdenum catalyst 2.6.7 which has become known as the Schrock catalyst (Figure 2. 5). It is highly active as it reacts with both terminal and internal alkenes and for the first time it made the formation of tri and tetra-substituted alkenes possible. It is also capable of performing ROMP reactions with low strain monomers and it can ring-close sterically demanding and electron deficient substrates. 15,16,17 It should be noted that the synthesis and handling of this catalyst require an inert atmosphere and thoroughly purified, dried and degassed solvents and reagents. This is due to the fact that the metal center is quite oxophilic and is thus oxygen and moisture sensitive. This fact also renders this catalyst less tolerant of many functional groups. For example, the Schrock catalyst is incompatible with protons on heteroatoms (carboxylic acids, alcohols, thiols etc.) and aldehydes. 18  Figure 2. 5 Common Metathesis Catalysts  N H 3C  CF3  F 3C  O  F 3C H3 C  PCy3 Cl Ru  Mo Cl  O Ph  Ph PCy3  Cl  PCy3 Cl Ru  Ph  PCy3  Ph  CF3  Schrock's Catalyst (2.6.7)  Grubbs 1st Generation Catalyst (2.6.4)  Grubbs 1st Generation Catalyst (2.6.9)  Following the invention of the Shrock catalyst, Professor Grubbs discovered two ruthenium based catalysts 2.6.4 and 2.6.9 depicted in Figure 2. 5. These catalysts exhibit much better functional group tolerance and they can also be used in the presence of water and oxygen. Unfortunately, these catalysts are not as active as the Schrock catalyst in that they cannot be used to form trisubstituted or tetrasubstituted olefins. Despite the reduced activity, the Grubbs first generation catalysts have found wide application. They have been used in ROMP and ADMET reactions to make polymers. The catalysts have also been used in RCM reactions to make small, medium and large ring systems which are commonly found in a wide variety of natural products.  48  Chapter 2 A First Generation Approach Towards Halichlorine…  Finally, the Grubbs first generation catalysts have been used in CM reactions of unhindered terminal olefins with symmetrically disubstituted internal olefins. The next breakthrough in catalyst design was reported by the Grubbs group in 1999 with the development of complex 2.6.10 (Figure 2. 6). 19 This complex has one of the tricyclophexylphosphine ligands found in the first generation Grubbs catalyst 2.6.4 replaced by an N-heterocyclic carbene (NHC) ligand. Phosphine dissociation was deemed to be an important part of the catalytic cycle. Therefore placing a strong σ-donor trans to the phosphine should aid in this dissociation process and lead to higher activity. As well, NHC’s are both strong electron donors and sterically bulky. Because of this NHC ligands should stabilize electron-deficient intermediates and promote olefin metathesis. Catalyst 2.6.10 turned out to be more active than its first generation predecessors although the reasons for the enhanced activity were later found to be different than those stated above. The high activity of ruthenium based metathesis catalysts with NHC ligands appears to be due to their improved selectivity for binding π-acidic olefins in the presence of free phosphine. This will be discussed in more detail below.  Figure 2. 6 Examples of Highly Active Second Generation Ruthenium Metathesis Catalysts N  N  Alk N  N Alk  Cl  N  Cl  Ru  Cl  Ru  Ph  Cl  N  Ph  Cl  PCy3 2.6.10  Cl  PCy3 2.6.11  N  PCy3 2.6.12  N  N  N Cl  Cl Ru  Ru Ph  Cl PCy3 2.6.8  Ph  Ru  Cl  O iPr 2.6.13  Concurrent with the work done in the Grubbs group, other groups were developing metathesis catalysts with modifications to the carbene ligands. 20 N-Heterocyclic carbenes with aryl groups on nitrogen were found to be more active than N-heterocyclic carbenes with alkyl groups on the nitrogen such as complexes 2.6.11 developed by Herrmann.20a,b Catalyst 2.6.12, developed by Nolan and Fürstner has the same N-heterocyclic carbene but has a phenyl-indenyl  49  Chapter 2 A First Generation Approach Towards Halichlorine… 20c,d  benzylidene ligand in place of the usual benzylidene.  This catalyst exhibited similar activity  to catalyst 2.6.10. Shortly after the initial report about catalyst 2.6.10, the Grubbs group reported the discovery of catalyst 2.6.8 which contains an N-heterocyclic carbene with a saturated backbone. This catalyst turned out to be even more active than catalyst 2.6.10. This was originally attributed to the enhanced σ-basicity of this N-heterocyclic carbene. Catalyst 2.6.13, developed by the Hoveyda group has an o-isopropoxybenzylidene ligand in place of the benzylidene and it displays a similar reactivity to catalyst 2.6.8. Collectively, the catalysts found in Figure 2.6 are referred to as the second generation metathesis catalysts. The second generation metathesis catalysts combine the high reactivity found in the early transition metal metathesis catalysts with the functional group tolerance of the first generation ruthenium metathesis catalysts. In addition, these catalysts are air and moisture stable therefore they do not require rigourously dried and distilled solvents and reagents. The second generation catalysts have wide ranging utility in all known metathesis reactions. ROMP can be done on both low-strain substrates and sterically hindered substrates that contain trisubstituted olefins, which could not be done with the first generation catalysts. RCM could now be done with sterically demanding dienes to form tri- and tetra-substituted olefins. Catalyst 2.6.8 was also the first catalyst that could perform CM to give trisubstituted olefins. CM and RCM could also be done on olefins containing electron withdrawing groups such as acrylates. This was not possible with the first generation catalyst. Metathesis has also been extended to alkene-alkyne CM and RCM where the alkyne can be disubstituted. With the second generation catalysts RCM could now be done with catalyst loadings as low as 0.05 mol % and ROMP could be done with catalyst loadings as low as 0.0001 mol %. Reaction times were reduced from hours and days to minutes and hours for terminal and disubstitutes olefins. Tri- and tetrasubstituted double bonds that either could not be formed with the first generation catalysts or required days to reach completion, could now be done in a matter of hours. It should be pointed out that initiation of the second generation catalysts require longer times and/or elevated temperatures (80 °C for 2.6.12) compared with the first generation analogs. However, once initiated the second generation catalysts seemed to react much more quickly. Overall, it was clear that the activity of the second generation catalysts was far superior to the first generation catalysts.  Chapter 2 A First Generation Approach Towards Halichlorine…  2.5.4  50  Mechanistic Considerations While many mechanistic schemes have been proposed over the years, the one proposed  by Chauvin remains the most widely accepted mechanism. 21 This mechanism involves the interconversion of a metal alkylidene and an olefin through a series of [2+2] cycloadditions and cycloreversions (Scheme 2. 7). It should be pointed out that this general mechanism indicates nothing about the nature of other intermediates formed during the reaction sequence.  Scheme 2. 7 Chauvin’s Proposed Mechanism for Metal Catalyzed Alkene Metathesis R1  [M]  + R2  R3  R1 R2  [M] R3  R1 R2  +  [M] R3  Following a series of mechanistic studies the Grubbs group proposed the mechanism outlined in Scheme 2. 8. It should be noted that the geometry around the metal center for all of the intermediates presented in Scheme 2. 8 is not indicated since thus far nobody has been able to provide direct evidence for the geometries of any of the intermediates. 22 For both catalysts 2.6.4 and 2.6.8 the first step involves dissociation of the bound tricyclohexylphosphine (PCy3) to form a 14-electron intermediate 2.8.2 (this process is defined as initiation). Originally it was postulated that an associative mechanism could be involved but this process was ultimately found to play no role in ruthenium-based metathesis reactions. The intermediate 14-electron species can either be trapped by free PCy3 or it can bind the substrate and undergo metathesis via a series of [2+2] and retro [2+2] reactions. The high activity of 2.6.8 which had previously been attributed to its ability to promote phosphine dissociation instead appears to be due to its improved selectivity for binding π-acidic olefins in the presence of free phosphine. While 2.6.8 does not dissociate phosphine as efficiently as 2.6.4, a small amount of 14-electron species is capable of several turnovers before it is deactivated by the rebinding of PCy3. On the other hand, 2.6.4 initiates more rapidly but rebinding with PCy3 is competitive with olefin coordination. Consequently, the 14-electron species produced from 2.6.4 undergoes relatively few turnovers before being trapped by free PCy3.  51  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 8 Proposed Mechanism for Ruthenium Metathesis Catalysts L Cl Ru  R  + PCy3  PCy3  + R 1  R  Cl2.Ru  - PCy3  Cl  k2  L  k1  -  k-1  2.8.1 k2 +  R1  k-2  R1  Cl2.Ru R1 2.8.3 k3  L Cl2.Ru R1 2.8.3  R  R1 k-2  2.8.2  L  k-3 L  R  k3 k-3  Cl2.Ru  R  R1 2.8.4  Overall the second generation catalysts have reached a level of reactivity where most metathesis reactions are conceivably possible. Disubstituted, trisubstituted and tetrasubstituted olefins can be formed using metathesis catalysts. Metathesis reactions can now be done in matter of minutes and hours instead of hours and days. As well metathesis can be done in the presence of a wide variety of functional groups. In recent years a number of research groups have developed chiral catalysts that are capable of doing asymmetric ring closing metathesis of C2-symmetric dienes and asymmetric kinetic resolution of racemic dienes. 23 However, with all the success of the second generation catalysts some challenges still remain. One challenge, which remains elusive, is the problem of olefin stereoselectivity in CM and in macrocyclic RCM. While tetrasubstituted alkenes can be formed in RCM of small rings, CM to form tetrasubstituted olefins also has not been done. 2.5.5  Summary The metathesis reaction has become extremely useful in organic chemistry. It has been  used in laboratories around the world to help make useful starting materials and it has been incorporated in countless total syntheses. This area has evolved from using poorly defined, low activity catalysts to single component, very reactive, stable catalysts that are capable of making highly functionalized olefins. Undoubtedly, this reaction will be a useful one in synthetic organic chemistry for many years.  52  Chapter 2 A First Generation Approach Towards Halichlorine…  2.6  Formation of the A Ring of Halichlorine by Ring Closing Metathesis: Ring Closing Metathesis and Elaboration of Diene 2.1.6a Earlier in this chapter, the synthesis of dienes 2.1.6a and 2.1.6b was described in Scheme 2.  6. In the last section the evolution of the metathesis reaction was described in some detail. The ring closing metathesis reactions of 2.1.6a and 2.1.6b will now be presented in addition to the subsequent transformations. As the two alkenes found in diene 2.1.6a are both terminal alkenes a first generation Grubbs catalyst should be reactive enough to do the ring closing metathesis reaction. In fact, Holmes and co-workers had successfully used Grubbs first generation catalyst 2.6.4 to do this exact transformation. 24 This reaction was successfully duplicated with a slight increase in yield (88 %, 84 % lit.) (Equation 2. 4).  Equation 2. 4 Ring Closing Metathesis of 2.1.6a PCy3 Cl Ru H  Cl  Ph CH2Cl2 (88 %)  N O  H  PCy3  2.1.6a  2  N O 2.1.5a  After making the bicyclic lactam 2.1.5a two objectives were identified. The first objective was to elaborate the lactam to an alkenyl stannane moiety and the second objective was to functionalize the C2 carbon (Figure 2. 7). Figure 2. 7 Objectives for Bicyclic Lactam 2.1.5a H  H 2  Functionalize C2  N  R O  2.1.5a  Convert Lactam to Alkenyl Stannane  N SnMe3  It had been well established in the Dake group that alkenyl stannanes similar to 2.1.5a can be made from the corresponding alkenyl triflates.10 Unfortunately, using the established protocol to make the lactam derived enol triflate of 2.1.5a produced a complex mixture of  53  Chapter 2 A First Generation Approach Towards Halichlorine…  undetermined products (Equation 2. 5). This result was not completely surprising as lactam derived enol triflates are known to be unstable especially when the substrate contains no electron withdrawing substituents. 25  Equation 2. 5 Attempted Enol Triflate Formation of Bicyclic Lactam 2.1.5a H  i) KHMDS, THF, -78 oC ii) Cl  N O 2.1.5a  N  complex mixture  N(Tf) 2 o  THF, -78 C to 0 oC  As halichlorine requires a carbonyl function at C2 an attempt was made to introduce functionality at the corresponding position of bicycle 2.1.5a. A possible method to achieve this goal would be to isomerize the alkene of 2.1.5a to enamide 2.9.1 (Scheme 2. 9) and to then introduce the carbonyl moiety by way of a Vilsmeir-Haack carbonylation reaction. It was also hoped that the presence of the α,β-unsaturated carbonyl function would impart some stability to the lactam-derived enol triflate anticipated in the next step. The isomerization reaction was accomplished by heating alkene 2.1.5a with triethylamine and palladium on carbon in THF to give compound 2.9.1. 26 Unfortunately the Vilsmeir-Haack carbonylation reaction failed to give the desired product. This failure was attributed to the fact that the enamide function of 2.9.1 is less nucleophilic than an enamine.  Scheme 2. 9 Attempted Elaboration of 2.1.5a O  Cl N  i) O  H  H NEt3, Pd/C 2  N O 2.1.5a  THF, 120 oC, 12 h (85 %)  O , PPh3, THF, reflux, 1h  N O  H , reflux, 1h ii) Me2N iii) then add 2.9.1, reflux, 1h iv) - THF, + H2O, reflux, 12 h  recovered starting material  2.9.1  An alternative way to introduce the carbonyl group would be to first exchange the hydrogen at C2 of 2.9.1 for bromine and to use a palladium catalyzed carbonylation reaction. When 2.9.1 was exposed to bromination conditions the desired alkenyl bromide 2.9.2 could be  54  Chapter 2 A First Generation Approach Towards Halichlorine…  obtained however the yields were generally low (≤ 25 %) (Equation 2. 6). In addition the yields were not reproducible and bromide 2.9.2 turned out to be quite unstable.  Equation 2. 6 Bromination of Enamide 2.9.1 Br2 or  H  O  Br N  CH2Cl2  N  -78 oC to rt (0 % to 25 %)  O 2.9.1  H  O  unstable  N  Br  O 2.9.2  Attempts to introduce functionality at C2 of bicycles 2.1.5a and 2.9.1 were unsuccessful. Therefore the ring closing metathesis reaction of diene 2.1.6b, to form bicyclic lactam 2.1.5b, was investigated next. Bicyclic lactam 2.1.5b contains the appropriate functionality at C2 required for the synthesis of halichlorine. 2.7  Formation of the A Ring of Halichlorine by Ring Closing Metathesis: Selection and Synthesis of a Second Generation Metathesis Catalyst In order to test the ring closing metathesis reaction of diene 2.1.6b, a second generation  metathesis catalyst was required. This was because one of the alkenes in diene 2.1.6b is an acrylate and because the desired product alkene is trisubstituted. The catalyst that was used to do the required transformation is a hybrid between the Grubbs second generation catalyst 2.6.8 and the Nolan/Fürstner catalyst 2.6.12 (Figure 2. 8). This was not done intentionally but turned out to be quite useful for us. This statement requires further explanation.  Figure 2. 8 Possible Ring Closing Metathesis Catalysts N  N  N  N Cl  Cl Ph PCy3 2.6.8 Grubbs 2nd Generation Catalyst  Cl  N Cl  Ph  Ru  Ru Cl  N  Ru Cl  PCy3 2.6.12 Nolan/Furstner Catalyst  PCy3 2.9.3 Our Hybrid Catalyst  Ph  Chapter 2 A First Generation Approach Towards Halichlorine…  55  To put things in perspective, this project was developed in early 2001. At that time the second generation catalysts had only recently been developed. In early 1999 Professor Grubbs reported catalyst 2.6.10 which contains the N-heterocyclic carbene (NHC) with the unsaturated backbone (Figure 2. 9). While this catalyst was found to be significantly more reactive than the first generation Grubbs catalysts no examples of RCM with acrylate alkenes were presented in the literature. Later in 1999 the Grubbs research group introduced catalyst 2.6.8 that contains the NHC with the saturated backbone. This complex was found to be more reactive than 2.6.10 and was able to do the same reactions as 2.6.10 but faster. Again, no examples of metathesis reactions with acrylate derivatives were given. It wasn’t until 2000 when Nolan and Fürstner reported the first comprehensive study on the formation of cyclic acrylate derivatives via ring closing metathesis. 27 Nolan and Fürstner used catalyst 2.6.12 to do these transformations. This catalyst was similar to the Grubbs second generation catalyst 2.6.10 except the benzylidene ligand was replaced with a phenylindenylidene ligand. They also reported that catalyst 2.6.12 exhibited a similar reactivity to catalyst 2.6.10 and that catalyst 2.6.12 was so stable that it could be purified by column chromatography. The reactivity of this catalyst towards acrylate derivatives, combined with the fact that it could be purified by column chromatography, made this catalyst attractive for use in our synthesis. In order to make this catalyst both the phenylindenylidene ligand and the NHC ligand would have to be synthesized.  Figure 2. 9 Second Generation Grubbs-type Catalysts N  N  N  N Cl  Cl Ru  Ru Ph  Cl  PCy3 2.6.8  PCy3 2.6.10 N  Ph  Cl  N  N Cl  N Cl  Ph  Ph  Ru  Ru Cl  Cl PCy3 2.6.12  PCy3 2.9.3  The phenylindenyl ligand comes from diphenyl propargyl alcohol, which can be made from trimethylsilyl acetylene and benzophenone (Scheme 2. 10). 28 Deprotonation of  56  Chapter 2 A First Generation Approach Towards Halichlorine…  trimethylsilyl acetylene (2.10.1) followed by addition to benzophenone (2.10.2) provided propargyl alcohol 2.10.3. Treatment with potassium carbonate in methanol removed the trimethylsilyl group to give diphenyl propargyl alcohol (2.10.4).  Scheme 2. 10 Synthesis of Diphenyl Propargyl Alcohol 2.10.4 TMS  H  i) BuLi ii) O Ph Ph 2.10.2  2.10.1  OH TMS  K2CO3, MeOH  OH H Ph Ph  Ph Ph 2.10.3  2.10.4  In order to make the desired catalyst 2.6.12, the NHC ligand had to be synthesized. It was during the synthesis of the NHC ligand that the mistake was made. Catalyst 2.6.12 contains the NHC ligand with the unsaturated backbone. The synthesis of the NHC ligand with the saturated backbone was completed when the mistake was discovered. Because catalyst 2.6.8 is more active than catalyst 2.6.10, we reasoned that catalyst 2.9.3 should be more active than catalyst 2.6.12. Therefore we decided to proceed with the synthesis of catalyst 2.9.3 as this catalyst should be able to promote the desired ring closing metathesis reaction. The synthesis of the unsaturated N-heterocyclic carbene ligand is described in Scheme 2. 11. 29 Glyoxal was stirred with 2,4,6-trimethylaniline to give the bis imine 2.11.3. Reduction with sodiumcyanoborohydride under acidic conditions gave N-(2-(mesitylamino)ethyl)-2,4,6trimethylbenzenamine (not shown). Subsequent treatment with ammonium tetrafluoroborate and triethylorthoformate gave imidazolium salt 2.11.4. 30  Scheme 2. 11 Synthesis of the N-Heterocyclic Carbene Ligand O  O  H H 2.11.1  +  MeOH rt (81 %)  H  H  N  N  MeOH, HCl (99 %) 2) NH4BF4, HC(OEt)3  NH2 2.11.2  1) NaCNBH3  +  N  N -  BF4  (44 %) 2.11.3  2.11.4  After making the requisite ligands, catalyst 2.9.3 was synthesized. This was done starting from ruthenium (III) chloride·nhydrate using the established protocols (Scheme 2. 12).  57  Chapter 2 A First Generation Approach Towards Halichlorine…  Hence, heating ruthenium (III) chloride·nhydrate (2.12.1) with triphenylphosphine in methanol gave tris(triphenylphosphine) ruthenium (II) chloride (2.12.2). 31 Exposure to diphenyl propargyl alcohol (2.10.4) resulted in the removal of one of the triphenylphosphine ligands and the addition of the phenylindenylidene ligand. 32 Subsequent treatment with tricyclohexylphosphine exchanged the remaining two triphenylphosphine ligands with tricyclohexylphosphine. Finally, tetrafluoroborate salt 2.11.4 was treated with potassium tert-butoxide and then mixed with metal complex 2.12.4.29a This resulted in the formation of the desired catalyst 2.9.3, which was purified by column chromatography to give a deep red solid. Unfortunately the yields for this reaction were quite low (~ 21 %). It was decided that the synthesis of catalyst 2.9.3 was not an important objective for us and therefore the optimization of this reaction was not explored.  Scheme 2. 12 Synthesis of the Second Generation Catalyst 2.9.3 OH H RuCl3.nH2O  PPh3, MeOH reflux (95 %)  2.12.1  Ru  PPh3  THF, reflux (67 %)  Ph  Ru Cl PPh3 2.12.3  i)  Mes N  +  N Mes BF4 2.11.4  Ph  KO Bu, THF ii) then add to a solution of 2.12.4 in PhH reflux (21 %)  N  Cl  2.12.4  N Cl  t  PCy3  2.8  Cl  Ru PPh3  2.12.2 PCy3 Cl  PCy3, CH2Cl2 (66 %)  Cl  PPh3 Cl  Ph Ph 2.10.4  PPh3  Ph  Ru Cl PCy3 2.9.3  Ring Closing Metathesis and Attempted Elaboration of Diene 2.1.5b After making metathesis catalyst 2.9.3 the ring closing metathesis reaction was attempted.  Heating the substrate 2.1.6b at 80 °C in toluene in the presence of 5 mol % of catalyst 2.9.3 resulted in complete conversion to the desired bicyclic lactam 2.1.5b after only twenty minutes (Scheme 2. 13). It should be noted that in the Nolan/Fürstner study on ring closing metathesis reactions of acrylate derivatives, similar transformations that used catalyst 2.6.12 required between two and twenty four hours to reach completion. Clearly, catalyst 2.9.3 was significantly more reactive. Unfortunately, catalyst 2.9.3 was not studied in any detail as the focus of my project was the total synthesis of halichlorine and not the development metathesis catalysts. It is  58  Chapter 2 A First Generation Approach Towards Halichlorine…  worth noting that this catalyst seems to be highly active. It is stable to column chromatography and it can be stored on the bench indefinitely with no noticeable loss in activity. 33  Scheme 2. 13 Ring Closing Metathesis and Attempted Elaboration of 2.1.6b H N  EtO 2C  O  2.9.3 PhCH3, 80 oC 20 min (quantitative)  H EtO  i) KHMDS, THF, -78 oC ii)Cl  N O  2.1.6b  O 2.1.5b  N  complex mixture  N(Tf) 2 o  , THF, -78 C to 0 oC  NEt3, Pd/C THF, 120 oC (4 %) H EtO  N O  O 2.13.1  After successfully closing the A ring of halichlorine, the next goal was to form the C ring. This was intended to be done via the previously mentioned acid catalyzed semipinacol rearrangement of allylic cyclobutanol 2.1.2. Previous work in the Dake group had shown that similar compounds can be made from alkenyl stannanes from lactam derived enol triflates. Therefore, lactam 2.1.5b had to be converted to the corresponding enol triflate. Unfortunately when 2.1.5b was treated with the standard enol triflate-forming conditions a complex mixture of products resulted (see Scheme 2. 13 for reaction details). This result was not entirely unexpected as bicyclic lactam 2.1.5b has more than one site which could be deprotonated with potassium hexamethyldisilazide (KHMDS). In addition to the site adjacent to the lactam carbonyl there are also protons on the carbon center located γ to the ester carbonyl which might be deprotonated by KHMDS, resulting in the formation of an extended ester enolate ion. To circumvent this problem an attempt was made to isomerize the alkene of 2.1.5b to give the corresponding vinylogous carbamate 2.13.1, similar to the conversion of 2.1.5a to 2.9.1 (see Scheme 2. 9 above for this reaction). This isomerization should eliminate the second deprotonation site mentioned above and provide a compound that when treated with KHMDS should deprotonate the proton adjacent to the lactam carbonyl and result in the formation of the desired enol triflate. However, in this instance the isomerization conditions did not prove successful as only a 4 % yield of  59  Chapter 2 A First Generation Approach Towards Halichlorine… -1  2.13.1 was obtained. Some basic modeling studies indicated that 2.1.5b is ~ 5 kcal·mol lower in energy than 2.13.1 therefore no further attempts to isomerize the alkene were attempted. At this stage one might imagine attempting to elaborate bicycle 2.1.5b through a sequence involving reduction of the ester, protection of the resultant alcohol and then enoltriflate formation. However, there were concerns that the ester and the lactam functionalities would display similar reactivities and that no selectivity would be observed in the reduction step. As well, concurrent to the work described above, progress towards the tricyclic core of halichlorine via Path B was starting to show some promising results. Therefore, Path A was abandoned in favour of Path B. It should be noted that the work done in Path A did produce some promising results. First of all a highly efficient, stable ring closing metathesis catalyst was made. Secondly, the ring closing metathesis reaction to form bicycle 2.1.6b appears to be a viable way to close the A ring of halichlorine. It was hoped that this reaction could also be used in Path B. 2.9 2.9.1  Progress Towards the Tricyclic Core of Halichlorine Via Path B Introduction As mentioned above, work towards the tricyclic core of halichlorine via Path B was  being done simultaneously to the work done for Path A. While both pathways incorporate a ring closing metathesis reaction, Path B differs in that the key semipinacol rearrangement reaction would be attempted prior to the ring closing metathesis reaction (Scheme 2. 14). It was believed that the substrate required for the semipinacol rearrangement could be made from the known alkenyl stannane 2.1.11.2  Scheme 2. 14 Retrosynthetic Analysis for Halichlorine via Path B H  H A RO2C  N  Path B  B C  O 2.1.1 H EtO Ts  O  N  +  O 2.1.8 H EtO Ts  2.1.4  N  Ts  O 2.1.9 O  N  H SnMe3  OH 2.1.10  N  RO2C  H  EtO  2.1.11  N O 2.1.7  60  Chapter 2 A First Generation Approach Towards Halichlorine…  2.9.2  Synthesis of Alkenyl Stannane 2.1.11 Alkenyl stannane 2.1.11 was previously synthesized from glutarimide (2.1.7).2 A brief  description of this synthesis follows (Scheme 2. 15). As was seen in Scheme 2. 6, 6-ethoxypiperidinone 2.6.1 could be made from glutarimide (2.1.7) in a one-pot reduction/acidification procedure. The lactam nitrogen was then protected as its p-toluenesulfonyl derivative 2.15.1. Conversion to the enol triflate 2.15.2 was accomplished by treatment with potassium hexamethyldisilazide and N-(5-chloropyridin-2-yl)-1,1,1-trifluoro-N[(trifluoromethyl)sulfonyl]methanesulfonamide (Comins reagent). 34 Stirring enol triflate 2.15.2 with hexamethyldistannane in the presence of a palladium (0) source gave the desired alkenyl stannane 2.1.11. The yields for all the steps in this sequence are comparable or better than those reported in the literature with the exception of the conversion of 2.15.1 to 2.15.2. The yield for this step was lower than that reported in the literature (48 %, lit. 97 %) but this reaction was only attempted once.  Scheme 2. 15 Synthesis of Alkenyl Stannane 2.1.11 O H  EtO  o  N O 2.1.7  i) NaBH4, EtOH, 0 C ii) pH=3 (72 %)  EtO i) KHMDS, THF, -78 oC ii) Cl N  N(SO2CF3)2  Ts  H  H  EtO  o  i) n-BuLi, THF, -78 C N O  ii) TsCl, THF, -78 oC to rt (53%)  Ts  EtO Me3SnSnMe3, Ph3As  N OSO 2CF3  Pd2dba3, THF (62 %)  2.15.2  N O  2.6.1  H  H  Ts  H  2.15.1  N SnMe3 2.1.11  THF, -78 oC to rt (48 %)  2.9.3  Synthesis of the Semipinacol Substrate 2.1.4 With alkenyl stannane 2.1.11 in hand, the semipinacol ring expansion substrate was then  made. A transmetallation reaction was done by treating alkenyl stannane 2.1.11 with n-butyllithium (Equation 2. 7). The intermediate alkenyl lithium (not shown) was then transmetallated again with magnesium (II) bromide to produce the corresponding Grignard  61  Chapter 2 A First Generation Approach Towards Halichlorine…  reagent. A subsequent carbonyl addition reaction with cyclobutanone (2.1.4) gave the desired allylic alcohol 2.1.10.  Equation 2. 7 Formation of Semipinacol Rearrangement Substrate 2.1.10 EtO Ts  H  i) MeLi, Et2O, -78 oC to 0 oC, 10 min o  EtO  ii) MgBr2, Et2O, -78 C, 30 min  N  iii) O  , Et2O, -78 oC, 2h; warm to rt overnight  Ts  H N  SnMe3  OH 2.1.4  (76 %)  2.1.11  2.9.4  2.1.10  Attempted Acid Catalyzed Semipinacol Rearrangement of 2.1.10 Unfortunately, exposure to acidic conditions did not result in a ring expansion reaction  (Scheme 2. 16). At low temperature the substrate did not react. Heating the substrate at a slightly elevated temperature resulted in decomposition of the starting material. It was speculated that the ethoxy-aminal moiety might be sensitive to the acidic conditions of the reaction and that this functional group might play a role in the decomposition of the substrate, possibly through an elimination-type mechanism. If elimination of ethanol is a problem, running the reaction with ethanol as the solvent might suppress the elimination reaction and result in the desired semipinacol rearrangement. However, when this reaction was attempted no semipinacol rearrangement products were isolated. Instead cyclobutylidene 2.16.1 was formed. Presumably this product is the result of a solvolysis reaction between ethanol and the substrate. From this outcome it was clear that this substrate was not suitable for an acid catalyzed semipinacol rearrangement. If a semipinacol rearrangement reaction were to be used to synthesize halichlorine, the synthetic plan would have to be modified.  62  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 16 Synthesis and Attempted Acid Catalyzed Semipinacol Rearrangement of 2.1.10 CSA, CH2Cl2, rt  H  EtO  N  Ts  No Reaction  CSA, CH2Cl2, reflux  Decomposition  OH o  2.1.10  CSA, EtOH, 60 C, 4h (49 %)  EtO Ts  H H OEt  N  2.16.1  2.9.5  Can the Ethoxy Group be Replaced by an Allyl Group? The ethoxy group located at the 6-position of the piperidine ring of ethoxy aminal 2.1.10  is not present in halichlorine. It was incorporated in the synthetic plan only as a means of further functionalization. Ultimately this plan involves conversion of the ethoxy group to an allyl substituent. Because the attempted acid promoted semipinacol rearrangement was unsuccessful with the ethoxy group present, a potential solution might be to attempt the ring expansion reaction on the 6-allyl derivative of ethoxy aminal 2.1.10. Therefore the 6-allyl derivative of ethoxy aminal 2.1.10, compound 2.16.2 was targeted as a possible ring expansion substrate (Table 2. 2). Table 2. 2 Attempted Allylation of 2.1.10 EtO  H SiMe 3  6  Ts  H  H  N OH  BF3.OEt2 (89 %)  2.1.10  6  Ts  +  N  Ts  N  OH 2.16.2  2.16.3  Entry Equiv. Allyltrimethylsilane Equiv. BF3·OEt2 Yield 2.22.1 Yield 2.22.2 1  3  2  0  90  2  1  2  0  40  Chapter 2 A First Generation Approach Towards Halichlorine…  63  It was believed that the most direct way to access the 6-allyl derivative of ethoxy aminal 2.1.10 was to attempt a nucleophilic addition reaction between allyltrimethylsilane and the iminium ion derived from ethoxy aminal 2.1.10 (Table 2. 2). It has been well established that alkoxy aminals can be ionized in the presence of a Lewis acid to produce intermediate iminium ions that can be trapped by nucleophiles such as allyltrimethylsilane. 35 When ethoxy aminal 2.1.10 was treated with three equivalents of allyltrimethylsilane in the presence of boron trifluoride·diethyl etherate, the bis-allylated compound 2.16.3 was isolated in 90 % yield as a single diastereomer (tentatively assigned as shown in Table 2. 2). When only one equivalent of allyltrimethylsilane was used, the bis-allylated product was isolated in 40 % yield in addition to unreacted 2.1.10. Unfortunately none of the desired mono-allylated product 2.16.2 was obtained. The apparent SN2′ reaction of the allylic alcohol seems to be competitive with ionization of the ethoxy aminal and nucleophilic attack on the iminium ion. These results were disappointing and led us to seek alternative methods to introduce the allyl group. Our next strategy involved installation of the allyl group at an even earlier stage in the synthesis. Recall from Path A where the synthesis of 6-allyl-piperidinone (2.6.2) was described (Scheme 2. 6). The nitrogen of this compound might be protected as its p-toluenesulfonyl derivative by using base and p-toluenesulfonyl chloride. If so, then 6-allyl-N-(toluene-4sulfonyl)-piperidinone (2.21.4) might be converted to the desired semipinacol rearrangement substrate by using standard reactions developed in the Dake group. Surprisingly, treatment of the 6-allyl lactam 2.6.2 with a variety of bases and p-toluenesulfonyl chloride resulted in only low yields (~ 5-27 %) of the desired compound 2.16.4 (Table 2. 3). It was speculated that the low yields might be a problem involving O-tosylation instead of the desired N-tosylation. Because the nitrogen of lactam 2.6.2 is quite hindered, O-tosylation might be favoured. The O-tosylate might get hydrolyzed in the workup to produce the starting material 2.6.2. In any event, these results were unsatisfying and our efforts were focused on other aspects of the project where some promising results were discovered. It should be noted that the idea of introducing the allyl group at an early stage in the synthesis was revisited and will be described later in this chapter.  64  Chapter 2 A First Generation Approach Towards Halichlorine…  Table 2. 3 Attempted N-Tosylation of Lactam 2.6.2 H  H  H  Base, TsCl Solvent  N Ts  O 2.6.2  N O  2.16.4  Recovered Starting  Entry  Base  Solvent  1  BuLi  THF  60  27  2  NaH  THF  80  10  3  KHMDS  THF  75  15  4  KOtBu  DMSO  30  5  2.9.6  Material (%)  Yield 2.16.4 (%)  Summary The semipinacol rearrangement substrate 2.1.10 was successfully synthesized however  the attempted acid catalyzed rearrangement reaction failed. This was probably due to the sensitivity of the ethoxy aminal functionality. Our efforts to introduce the allyl group prior to attempting the acid catalyzed ring expansion reaction also were unsuccessful. From these results it was evident that in order to move forward a new idea was required. We were still interested in using a semipinacol rearrangement to build the 6-azaspirocyclic ring system but using acid to do this was not a viable option. A possible solution to this problem would be to use electrophiles other than Bronsted acids to promote the semipinacol rearrangement reaction. Specifically, electrophilic halogen reagents were investigated.  Chapter 2 A First Generation Approach Towards Halichlorine…  65  2.10 Electrophilic Halogenation Reactions of Alkenes 2.10.1 Introduction In the last section it was demonstrated that semipinacol substrate 2.1.10 was unable to undergo a semipinacol rearrangement when a Bronsted acid was used as the electrophile. Alternative electrophilic reagents are known to promote semipinacol rearrangement reactions. While there are numerous electrophilic reagents that could be used, ultimately N-bromosuccinimide (NBS), an electrophilic bromine source, proved to be successful. Before describing the specific reactions that were done using N-bromosuccinimide, an account of electrophilic halogenation reactions of alkenes will be given. There are several electrophilic reagents that have been used in reactions with alkenes. Some examples include: chlorine, N-chlorosuccinimide, bromine, N-bromosuccinimide, iodine, N-iodosuccinimide, mercury (II) acetate, phenyl selenium chloride, ozone and borane. It is noteworthy that several of the reagents listed are electrophilic halogen sources. While the uses of all of these electrophiles could be discussed in detail, only the reactions between electrophilic halogen sources and alkenes will be discussed here. 2.10.2 Reactions of Alkenes with Diatomic Halogen Reagents The addition of diatomic halogens across double bonds is a reaction that dates back to some time in the late 1800’s to the early 1900’s. Upon reading the literature it is not clear who was the first to perform this type of reaction as nobody makes reference to a specific paper. However, it was found that bromine (Br2) or chlorine (Cl2) can add across double bonds to give dihalogenated products. For example, the reaction of bromine with cyclopentene results in the formation of racemic trans-1,2-dibromocyclopentane while none of the cis product is formed (Equation 2. 8).  Equation 2. 8 Stereospecific Reaction of Bromine with Cyclopentene Br2  Br  +  Br  Br Br racemic trans-1,2-dibromocyclopentane  Chapter 2 A First Generation Approach Towards Halichlorine…  66  2.10.3 Mechanism Over the years there has been some debate over the exact mechanism involved in these types of reactions. It has always been generally accepted that the reaction occurs in two steps: the first step involves attack of the π-electrons of the alkene onto bromine (or chlorine) resulting in the formation of a cationic intermediate with bromide (or chloride) as the counterion (Scheme 2. 17). 36 In the second step the resulting carbocation is attacked by bromide (or chloride) to give the observed product. In 1932 Robinson originally proposed that the cationic intermediate formed in the first step was a discrete carbocation. However, over time it became evident that most alkenes react to give predominantly the trans dihalogenation products. Most people believed that a discrete carbocation intermediate should result in mixtures of cis and trans products instead of the predominance of the trans products. In an attempt to explain the diastereoselectivity, Roberts and Kimbell proposed a new intermediate for the first step of the reaction. They proposed a 3-centred cationic intermediate involving both carbons of the alkene and the halogen (Scheme 2. 17). This intermediate would later be called a halonium ion. The halide would then attack the onium ion preferentially in an antiperiplanar manner. This mechanism leads to the predominant formation of the trans dihalogen compound and thus would account for the major product observed from the addition of bromine (or chlorine) to most alkenes.  Scheme 2. 17 Two Possible Mechanisms for the Addition of Halogens (X2) to Alkenes Robinson's Proposed Mechanism H Ph  H  Br2  Ph  H  H + Ph Br Ph  Br  -  H Ph  Br Br  Ph H  Roberts and Kimbell's Proposed Mechanism R  R  R  R  X2  R  X R +  R  R  X  -  X R R  R R X  onium ion  The name for a specific halonium ion reflects the type halogen present in the ion. For instance, if the ion contains a bromine atom it would be called a bromonium ion. Similar 3-centered cationic intermediates have also been proposed for mercury and selenium.  Chapter 2 A First Generation Approach Towards Halichlorine…  67  Collectively these types of ions are referred to as onium ion intermediates. Since Roberts and Kimbell originally proposed the idea of halonium ion intermediates there has been much indirect and direct evidence to support this claim. 37 It is now generally accepted that most cyclic alkenes react with electrophilic halogen reagents by forming halonium ion intermediates. It has also been suggested that some cyclic alkenes react through halogen alkene π-complexes rather than onium ions. The difference in these mechanisms seems to be subtle as both transition states usually result in the formation of same products, although some exceptions are known. 38 Certain acyclic alkenes have been found to react by halogen alkene π-complexes resulting in the formation of products with configurations different from those that would be predicted by the intermediacy of an onium ion intermediate. 39 These examples are beyond the scope of this discussion and will not be presented here. Fluorine (F2) is known to add to alkenes but fluorine is highly reactive and its reactions are difficult to control. Therefore only limited work has been done involving the addition of fluorine to alkenes. On the other hand, iodine (I2) does not undergo addition to alkenes. It is believed that iodine is capable of forming the cationic intermediate but that the second step is not favourable. The formation of the cationic intermediate is reversible and therefore the reaction favours the starting materials. A significant amount of work has been done on the addition of bromine (Br2) and chlorine (Cl2) to alkenes. In general it is known that bromine (or chlorine) adds to alkenes in a stereospecific manner to give predominantly the trans dihalogenation product. 2.10.4 Reactions of Simple Alkenes where the Electrophilic Halogen Source and the Nucleophile are Different So far the reaction of a simple symmetrical cyclic alkene with a diatomic halogen reagent has been presented. There are examples where the nucleophile in the second step is different from the electrophile. In addition, reactions can also be done with unsymmetrical alkenes. These reactions can be both stereoselective and regiospecific. An informative example that demonstrates this is the reaction of 1-methyl-1-cyclopentene (2.17.1) with bromine and water to form racemic trans-2-bromocyclopentanol (2.17.2 and 2.17.3) (Equation 2. 9). Note that the OH and the Br substituents are trans and the expected Markovnikov products are formed exclusively. 40  68  Chapter 2 A First Generation Approach Towards Halichlorine…  Equation 2. 9 Halohydrin Reaction OH Br2, H2O Br 2.17.2  2.17.1  OH  +  Br 2.17.3  In the above example 1-methyl-1-cyclopentene is not symmetrical. Consequently the bromonium ion intermediate is also asymmetric. There should be a larger partial positive charge (denoted by δ+ in Scheme 2. 18) on the more highly substituted carbon. Water will attack this carbon antiperiplanar to the bromonium ion leading to the formation of intermediate 2.18.2. Subsequent loss of a proton gives the observed product 2.17.2. No products resulting from attack of water on the least substituted carbon of bromonium ion intermediate 2.18.1 were observed.  Scheme 2. 18 The Addition of Bromine and Water to 1-Methyl-1-cylopentene Br  Cl Cl H  CH3 2.17.1  H  CH3  Br δ+ CH3  OH2  H  +  Br  OH2  2.18.1 2.18.2 - asymmetric bromonium ion - δ+ larger on the more highly substituted carbon - water attacks most substituted carbon  CH3 H  OH 2.17.3  2.10.5 Diastereoselective Reactions of Chiral Cyclic Alkenes So far the reactions that have been discussed have been reactions of simple achiral cycloalkenes with an electrophilic halogen and an external nucleophile. Similar reactions have also been performed on cyclic alkenes that contain stereogenic centers. Reactions of alkylsubstituted cycloalkenes have not been widely studied. However, of the few examples that have been done the reactions tend to be non-selective (Scheme 2. 19). The reactions of 4-substituted cyclohexenes with iodine and bismuth (III) acetate are non selective resulting in the formation of mixtures of stereo- and regio-isomeric trans-cyclohexane-1,2-iodoacetates. 41 Intererestingly, when 3-methylcyclohexene was exposed to the same reaction conditions 2.19.12 was formed as the major product (81 %). In this reaction the iodine adds to the face of the cyclohexene  69  Chapter 2 A First Generation Approach Towards Halichlorine…  opposite to the methyl substituent on the alkene carbon that is proximal to the methyl group. No rationale was provided for this outcome.  Scheme 2. 19 Addition Reactions with Substituted Cyclohexene Compounds  1 equiv Bi(OAc)3  4  3 equiv I2 dry AcOH (86 %) 2.19.1  I  +  I  OAc 2.19.2  +  +  OAc  OAc  OAc 2.19.3  I 2.19.5  I 2.19.4  CH3 1 equiv Bi(OAc)3  4  3 equiv I2 dry AcOH (91 %) 2.19.6  3  CH3  +  OAc  I  I  2.19.7  2.19.8  OAc I (81 %) 2.19.12  +  I  +  +  OAc  OAc I 2.19.13  I OAc 2.19.10  2.19.9 CH3  CH3  1 equiv Bi(OAc)3 3 equiv I2 dry AcOH  2.19.11  OAc  CH3  +  I OAc (10 %) 2.19.14  CH3  +  I OAc 2.19.15  2.10.6 Diastereoselective Reactions of Cycloalkenes that Contain Alcohols/Ethers Reactions of cycloalkenes that contain alcohols or ethers have also been studied. Specifically a significant amount of work has been done involving cyclic allylic alcohols. On the other hand, with the exception of highly functionalized cycloalkenes, most notably in the context of carbohydrate chemistry, the reactions of simple cycloalkenes with oxygenated substituents at other positions have not been widely studied. Reactions of alkenes that contain allylic heteroatoms such as oxygen or nitrogen often give products that have enhanced diastereoselectivity. A notable example is the asymmetric epoxidation of allylic alcohols. 42 High diastereoselectivities are also observed when allylic alcohols are treated with electrophilic halogen sources and nucleophiles (Scheme 2. 20). 43 A number of important observations should be made. As expected the halogen and the nucleophile have a trans relationship. This is consistent with an onium ion intermediate. In addition the halogen adds to the alkene on the face syn to the alcohol. Finally the halogen shows a preference for adding to the alkene carbon that is proximal to the alcohol bearing carbon.  Chapter 2 A First Generation Approach Towards Halichlorine…  70  Scheme 2. 20 Reactions of Cyclic Allylic Alcohols/Ethers OH  OH  NBS H2O DMSO  OH Br  OH 2.20.2 4  2.20.1  Br  +  OH 2.20.3 1  OH  OH  Br  HOBr  OH t-Bu  t-Bu  2.20.4  2.20.5  It is widely believed that the electrophile attacks the alkene syn to the alcohol because of stabilizing interactions between the oxygen lone pair and the onium ion (Scheme 2. 21).39d It should be noted that unless the cyclic allylic alcohol can be conformationally “locked”, such as in t-butyl cyclohexenyl or bicyclic substrates, the diastereoselectivities tend to be lower than their acyclic counterparts.  Scheme 2. 21 Reactions of Cyclic Allylic Alcohols O  H  H  O  E +  H lone pair stabilizes the onium ion  H  E  OH  H2O HO "syn" (major)  There are examples of reactions involving cyclic allylic alcohols/ethers where the electrophile adds to the alkene selectively on the face opposite to the alcohol/ether (Equation 2. 10). 44 In addition the halogen preferentially ends up on the alkene carbon that is distal to the alcohol-bearing carbon. Both of these trends are opposite to the observed results from the reactions of cyclic allylic alcohols/ethers with electrophilic halogen reagents and nucleophiles given above.  71  Chapter 2 A First Generation Approach Towards Halichlorine…  Equation 2. 10 Opposite Diastereoselectivity for Reactions of Cyclic Allylic Ethers SMe O  NBn  O I2  O NBn I  2.21.1  2.21.2  In this case there is an internal nucleophile which leads to the change in both regio- and diastereoselectivity. Formation of the “unstabilized” onium ion on the face opposite to the alcohol/ether is followed by intramolecular cyclization anti to the onium ion (Scheme 2. 22). The presence of the internal nucleophile makes this pathway a thermodynamic “sink”. So, while the syn onium ion may form, the absence of a nucleophile makes this pathway unproductive. Because onium ion formation is believed to be reversible, everything funnels back through the other onium ion leading to the cyclized product.  Scheme 2. 22 Reactions of Cyclic Allylic Alcohols with an Internal Nucleophile Nuc O + E  E  H  O Nuc  Nuc O H  Nuc +E  O H  X  Unproductive  2.10.7 Reactions of Glycals  Figure 2. 10 Glycal Numbering Scheme 5 6  4  1O  3 2  2.22.1  Glycals such as 2.22.1 are similar to compound 2.1.10 both structurally and electronically (Figure 2. 10). Therefore it would be useful to discuss the reactivity of these types of  72  Chapter 2 A First Generation Approach Towards Halichlorine…  compounds towards electrophilic halogen reagents. The numbering scheme depicted in Figure 2. 10 will be used throughout the following discussion. The reader should be aware that the numbering scheme used here is different from the one typically used to describe these systems. The typical numbering scheme is based on the carbohydrate numbering system in which the oxygen is not assigned a number. The numbering scheme used here includes oxygen in the numbering scheme and is designated the number 1. This numbering scheme is used so that comparisons can be easily made with substrates that appear later in this chapter. The following discussion will involve the reactions of unsubstituted glycols and 6-substituted-glycals. To the best of my knowledge, the reactions of glycals that only contain substituents at the 4-position or the 5-position of the pyran ring have not been studied. Before discussing the specific examples a brief discussion will be given on the different mechanisms through which glycals react with electrophilic halogen reagents.  Scheme 2. 23 Possible Transition States for the Reaction of Glycals with Electrophilic Halogen Reagents Halogen Alkene π-complex X X X2  O  O  O  - concerted - early transition state  X Nuc  Nuc Onium Ion X2 O  O  X  O  +  - may be concerted - later transition state  X Nuc  Nuc Halogen Oxacarbenium Ion X2 O  O  +  O  X Nuc  +  X  - step-wise X - late transition state  O Nuc  Unlike all carbon cycloalkenes glycals are known to react via different mechanisms.38b Along the reaction coordinate there is a continuum of transition states. In certain cases these reactions can proceed through halogen alkene π-complexes which are commonly referred to as  Chapter 2 A First Generation Approach Towards Halichlorine…  73  being relatively early along the reaction coordinate (Scheme 2. 23). There is a sub-group of halogen alkene π-complex transition states with some occurring earlier or later than others. Halogen π-complexes that react later along the reaction coordinate involve a significant amount of charge build-up in the transition state. Reactions that proceed through halogen alkene π-complex transition states are generally considered to be concerted. Alternatively, some reactions proceed through onium ion intermediates. Substrates that react via onium ion intermediates do so in a stepwise manner. Onium ion transition states occur later along the reaction coordinate compared with halogen π-complex transition states. Finally, other reactions can progress through a two-step sequence involving the initial formation of a halogen oxacarbenium ion. The formation of this intermediate is irreversible and these reactions are described as being even later along the reaction coordinate. Obviously each of these mechanisms has implications for the products that are formed in these reactions. Reactions of unsubstituted glycals typically react to give the trans products. A representative example is depicted in Equation 2. 11. 45 The product outcome is consistent with the reaction proceeding through an onium ion intermediate.  Equation 2. 11 Reaction of an Unsubstituted Glycal NIS HO O  CH2Cl2, 0 oC  O  I O  2.23.1  (37 %) 2.23.2  +  O  I O  (45 %) 2.23.3  A similar trend is also observed for substrates with substituents at the 6-position (Table 2. 4).38b In all cases the major product is the one in which the electrophile and the R group are cis and where the nucleophile is trans to both the R group and the electrophile. The diastereoselectivity is high for all of the examples with the exception of those run in methanol (entries 7 and 8). Because none of the reaction were completely diastereoselective some comments regarding the product outcomes and their mechanistic implications will be given.  74  Chapter 2 A First Generation Approach Towards Halichlorine…  Table 2. 4 Reactions of 6-Substituted 2,3-Dihydropyrans R1  Conditions  R1  O  O  1  E  +  Nuc 2.23.5  2.23.4  Entry  R1  R1  R1 O  E  Nuc 2.23.6  Reagents Solvent  CH2OBn ICl, NaN3 CH3CN ICl, NaN3 CH3CN  +  O  E  Nuc 2.23.7  Nuc  E  % 2.23.5 % 2.23.6 % 2.23.7  N3  I  87  13  0  N3  I  84  16  0  2  Ph  3  CH2OBn  Br2  DCE  Br  Br  95  0  5  4  Ph  Br2  DCE  Br  Br  86  0  14  5  CH2OBn  Bu4NBr3  DCE  Br  Br  95  0  5  6  Ph  Bu4NBr3  DCE  Br  Br  90  0  10  7  CH2OBn  Br2  MeOH  MeO Br  56  44  0  8  Ph  Br2  MeOH  MeO Br  62  38  0  The reactions listed in entries 1 and 2 from Table 2. 4 are believed to proceed through a two-step sequence involving an onium ion intermediate. 46 Azide ions are strong nucleophiles and they are known to react with carbocations under diffusion-control. 47 Therefore the product distribution is the result of kinetic control. Bromine additions to glycals in aprotic solvents (DCE) are believed to react through oxacarbenium ions (entries 3 and 4). 48 While the trans addition product is predominantly formed the open transition state results in the formation of small amounts of the cis product. Reactions of glycals with tribromide ion are known to occur through “late” bromine π-complexes.38a Hence, the trans addition occurs through a concerted process leading to the major product observed. However, tribromide ion is also in equilibrium with bromine (Br2) and bromide ion (Br-). Because bromine can react through a bromo oxacarbenium ion a small amount of the cis product is observed. Finally, when the bromination reaction is carried out in methanol the reaction is non-selective. However the reaction is completely selective for the trans addition of bromine and methanol across the double bond. The authors submit that these reactions occur through “early” bromine π-complexes. They also say that these reactions occur so fast that ionic species cannot exist in methanol. This leads to the non-selective formation of the two diastereomeric products observed.38b Obviously there are a number of factors which govern the reactions of glycals with electrophilic halogen reagents. These include: solvent effects, reagent effects, steric effects, and  Chapter 2 A First Generation Approach Towards Halichlorine…  75  electronic effects. In addition the product outcomes are dependant upon the position of the transition state along the reaction coordinate. However, in general it seems as though the major product is the one that contains a cis relationship between the electrophile and the R group and where the nucleophile is trans to both the R group and the electrophile. It should be noted that studies have been done involving the reactions of polysubstituted glycals where the diastereoselectivities are quite high. 49 The authors have attempted to rationalize the stereochemical outcomes of many of these reactions. Because of the highly functionalized nature of the substrates the factors that affect configuration become even more complex. These discussions, while interesting, are beyond the scope of this thesis and will not be mentioned here. Instead, the ring expansion reactions of allylic cycloalkanols that are promoted by electrophilic halogen reagents will be presented. These reactions are quite relevant to this discussion as compound 2.1.10, i.e. the ring expansion substrate, is an allylic cyclobutanol. 2.10.8 Electrophilic Halogen Promoted Ring Expansion Reactions of Cycloalkyl Allylic Alcohols The use of electrophilic halogens to promote the rearrangement of allylic cycloalkanols has not been widely studied. In fact at the time of preparation of this manuscript, only thirteen examples had been reported. A discussion of all of these contributions will be presented in the following pages. The first examples of electrophilic halogen promoted ring expansion reactions of allylic cycloalkanols were done by Johnson in 1964 (Scheme 2. 24). 50 This work involved the use of tert-butylhypochlorite to promote the ring expansion of a variety of allylic cycloalkanols. This reagent was used to promote the ring expansion of 4→5, 5→6 and 6→7 membered rings. They assumed that the reaction proceeds through the intermediacy of a chloronium ion. Johnson also attempted to promote these ring expansions with bromine however only small amounts of ring expansion products were generated. They claimed that bromine promotes ring expansions to a lesser extent than chlorine because the bromonium ion is more stable than the analogous chloronium ion.  Chapter 2 A First Generation Approach Towards Halichlorine…  76  Scheme 2. 24 The First Electrophilic Halogen Promoted Ring Expansion Reactions OH  OCl  O Cl ( )n  ( )n  2.24.2a (n=0, 68 %)1 2.24.2b (n=1, 38 %) 2.24.2c (n=2, 34 %)  2.24.1a (n=0) 2.24.1b (n=1) 2.24.1c (n=2)  O HO  2.24.2d2  2.24.1d 1 2  Cl  OCl  Isolated as its semicarbazone Recrystallized yield not reported although the crude yield was over 100 %  A few years later Wasserman and co-workers were able to promote the expansion of allylic cyclopropanols to the corresponding cyclobutanones (Scheme 2. 25). 51 They also showed that propargyl cyclopropanol 2.25.5 could be expanded to give 2-chloromethylenecyclobutanone 2.25.6. In a follow-up paper they successfully used bromine to promote the ring expansion of cyclopropanol 2.25.1 to cylobutanone 2.25.2b. 52  Scheme 2. 25 Ring Expansions of Cyclopropanols HO  OCl or Br2  X  2.25.2a (X=Cl, 81 %) 2.25.2b (X=Br, 45 %)  2.25.1 HO  O  OCl  O  Cl  (80 %) 2.25.3 HO  2.25.4 OCl  O  Cl  (60 %) 2.25.5  2.25.6  The Trost group was the first to provide some insight into the mechanism of electrophilic bromine promoted ring expansion reactions of allylic cyclopropanols (Scheme 2. 26). 53 They used dioxane·Br2 complex to convert silylether 2.26.1 into a 1:1 mixture of cyclobutanones 2.26.2a and 2.26.2b. In a similar fashion they were able to transform allylic cyclopropanol 2.26.5 into a 1:1 mixture of cyclobutanones 2.26.6a and 2.26.6b. Interestingly, this specific  Chapter 2 A First Generation Approach Towards Halichlorine…  77  example indicates that the reaction is quite chemoselective as only the allylic cyclopropanol reacts and not the isolated double bond. This was attributed to a directing effect by the alcohol functional group. Trost submitted that open carbocations may be responsible for the lack of diastereocontrol in both of these examples. He also argued that a bridged intermediate should force migration trans to the electrophile and lead to enhanced diastereoselectivity. To test this presumption they ran the analogous epoxidation reactions of 2.26.3 and 2.26.5 (note that the epoxidation reaction was run on the free alcohol of 2.26.1 instead of the silyl ether). Ring expansion of the corresponding epoxides, which necessitate bridged intermediates, were much more selective giving almost exclusive formation of cyclobutanone 2.26.4 and an 8:1 mixture of cyclobutanones 2.26.7a and 2.26.7b. The major product from the epoxidation reactions was the one where the migrating group was trans to what was the epoxide oxygen. The results from the epoxidation reactions seem to support Trost’s claim that the bromine promoted reactions were going through an open carbocation intermediate. The ring expansion reactions of other ring sizes were not tested. As a result the claim that open carbocations are involved in the ring expansion reactions involving other ring sizes could not be made.  Scheme 2. 26 Ring Expansion Reactions with Bromine and tert-Butyl Hydroperoxide O OTMS  O  dioxane.Br2 Br 2.26.2a  2.26.1  Br 2.26.2b (a:b = 1:1)  O OOH OH  cat. VO(acac)2  OH 2.26.4  2.26.3 H  HO  H  Br O  2.26.5 H  HO  H 2.26.5  H  H dioxane.Br2  Br  +O H  2.26.6a (a:b = 1:1)  cat. VO(acac)2  HO O  2.26.6b H  H  OOH  H  HO  +O H  H  2.26.7a 2.26.7b (a:b = 8:1)  78  Chapter 2 A First Generation Approach Towards Halichlorine…  There were no further reports of electrophilic halogen promoted rearrangement reactions of allylic cycloalkanols until the year 2000. In efforts to access new camphor derivatives to use as chiral starting materials in asymmetric synthesis, Cerero and Martínez used an N-bromosuccinimide promoted rearrangement reaction to make 10-bromocamphor and 10-bromofenchone (Scheme 2. 27). 54,55 Treatment of (1R)-camphor derivative 2.27.1 with N-bromosuccinimide leads to an intermediate carbocation 2.27.2. This rearranges to give a new carbocation 2.27.3 which after loss of H+ with concomitant C-O π-bond formation gives (1R)-10-bromocamphor (2.27.4). A similar transformation was used to convert (1R)-fenchone into (1S)-10-bromofenchone (2.27.6).  Scheme 2. 27 Synthesis of Chiral Bromo-Camphor Derivatives OH  O  Br N  O  OH Br  +  CH2Cl2 (96 %)  2.27.2  OH  O  CH2Cl2 (96 %) 2.27.5  OH Br  - H+  O  Br  Br  2.27.1 Br N  +  2.27.3  O  2.27.4  O  Br  O  2.27.6  An interesting project in the Paquette laboratories has involved the synthesis of spirocyclic furans and pyrans via acid promoted semipinacol rearrangement reactions of substrates such as allylic cyclobutanol 2.28.1 (Scheme 2. 28). Treatment of allylic cyclobutanol 2.28.1 with acid forms an intermediate oxonium ion 2.28.2 that undergoes a rearrangement reaction, C-O π-bond formation and loss of a proton to give the 1-oxaspirocycle 2.28.3. 56 In an isolated experiment the Paquette group showed that a similar transformation could be carried out using bromine as the electrophile. Exposing allylic cyclobutanol 2.28.1 to N-bromosuccinimide in a 1:1 mixture of isopropanol and propyleneoxide at low temperature resulted in the formation of 6-oxaspirocyclopentanone 2.28.5 as the only product. Because of the diastereoselectivity observed in the reaction, the Paquette group reasoned that the reaction must involve a bromonium ion intermediate 2.28.4. This was the only example reported by the Paquette group.  79  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 28 Paquette’s NBS Promoted Ring Expansion via an Oxonium Ion O  OH  H O  H+ (45-87 %)  O  +  O O  2.28.1  2.28.2  O O  OH  Br N  O  H O  i-PrOH, O  +  Br O  Br O O  -78 oC to rt (96 %) 2.28.1  2.28.3  2.28.4  2.28.5  2.10.9 Summary The reactions between cyclic alkenes and electrophilic halogen reagents have been presented. These reactions tend to show high degrees of selectivity across a wide range of substrates. The reactions are generally thought to proceed though onium ion intermediates although halogen π-complexes may also be involved. The products that are formed are usually those resulting from trans addition across the double bonds, The diastereoselectivities observed in the reactions of cyclic alkenes that contain stereogenic centers can be quite good and in many cases the configuration of the products can be predicted. However, a number of factors can contribute to the diastereoselectivities observed. These include: solvent effects, reagent effects, steric effects, and electronic effects. Electrophilic halogen reagents have also been used to promote the ring expansion of some allylic cyclobutanols, although there have only been a limited number of examples reported. As such there remains some debate as to the exact nature of the mechanism involved. With the exception of Paquette’s example the reactions studied thus far have not shown any level of diastereoselectivity. It would be useful if these types of reactions could be done across a wide variety of substrates and at the same time impart high levels of diastereoselectivity.  80  Chapter 2 A First Generation Approach Towards Halichlorine…  2.11 An Electrophilic Bromine Solution to the Semipinacol Rearrangement Reaction 2.11.1 Attempted Semipinacol Rearrangement Reactions with Allylic Alcohol 2.1.10 Recall that the acid catalyzed rearrangement of allylic cyclobutanol 2.1.10 was unsuccessful. In the last section semipinacol rearrangements promoted by electrophilic halogen reagents were presented. In particular the conditions used by Paquette to promote the ring expansion of allylic cyclobutanol 2.28.1 to spirocyclobutanone 2.28.5 (Scheme 2. 28) were quite interesting as only one diastereomeric product was formed. The semipinacol rearrangement of allylic alcohol 2.1.10 was attempted using these conditions in addition to another test electrophile, phenyl selenium chloride (Scheme 2. 29). Unfortunately phenyl selenium chloride produced a complex mixture of at least three compounds that were inseparable by chromatography. In contrast, treatment of allylic alcohol 2.1.10 with 1.2 equivalents of N-bromosuccinimide in a 1:1 mixture of isopropanol and propylene oxide at -78 °C resulted in the formation of 6-azaspirocycle 2.29.1 as a single diastereomer. The reaction was complete within three hours.  Scheme 2. 29 Successful Semipinacol Rearrangement Promoted by N-Bromosuccinimide H  EtO  N  Ts  PhSeCl  Complex Mixture  OH 2.1.10  EtO Ts  O  H  PrOH,  i  N OH  2.1.10  Br N  ο  O O  -78 C to rt (quantitative)  EtO Ts O  H  H Br N  Br  O  O N S O O  2.29.1  ent 2.29.1  H  Chapter 2 A First Generation Approach Towards Halichlorine…  81  2.11.2 Determination of the Stereochemical Configuration of 2.29.1 The formation of spirocyclopentanone 2.29.1 was confirmed by several pieces of spectral information. The presence of a carbonyl was supported by: the presence of a strong signal at 1740 cm-1 in the infrared (IR) spectrum and the presence of a chemical shift at δ = 214.3 in the 13  C NMR spectrum. The chemical shifts corresponding to the alkene proton (δ = 5.75) and the  tertiary alcohol proton (δ = 4.45) of 2.1.10 were both absent in 1H NMR spectrum of the product. Low resonance electrospray ionization (ESI) mass spectrometry gave mass peaks of 451 and 453 that correspond to the expected parent masses that contain either 79Br and 81Br, the two most abundant bromine isotopes, plus sodium. 57 The stereochemical assignments were made easier by the presence of the bromine in the product. The chemical shift for the proton on the brominebearing carbon was easily identifiable in the 1H NMR spectrum at δ 4.61 and the associated coupling constants (J = 11.6, 4.3 Hz) are consistent with this proton occupying an axial or pseudoaxial orientation. It follows that the bromine is equatorial or pseudoequatorial. Because the migration was predicted to proceed in an antiperiplanar manner, the configuration of the spirocenter could be assigned. Ultimately, the structure and relative configuration of the compound was confirmed by single crystal X-ray analysis. The ORTEP representation shows that our stereochemical assignments were correct. 2.11.3 Mechanistic Rationale to Explain the Formation of Spirocyclopentanone 2.29.1 To explain the configuration the mechanism depicted in Figure 2. 11 can be used. The ethoxy group is situated in a pseudoaxial orientation so that A1,3 strain with the p-toluenesulfonyl group on the nitrogen is minimized. This inference is supported by the small coupling constants (J = 3.7, 2.7 Hz) for the 1H NMR signal corresponding to the pseudoequatorial methine proton at the 6-position of the piperidine ring. N-Bromosuccinimide approaches the alkene on the sterically less hindered face of the more stable half-chair conformation. Formation of a bromonium ion or a bromine π-complex triggers an alkyl shift, which after loss of a proton results in the formation of the cyclopentanone. The results are consistent with antiperiplanar migration of the alkyl group where the migrating group and the bromine end up trans.  82  Chapter 2 A First Generation Approach Towards Halichlorine…  Figure 2. 11 Proposed Mechanism To Explain Configuration H  EtO  EtO H Ts N  6  Ts  N  EtO Ts O  OH  OH  H N  Br  +  2.10.10  Br - EtO axial due to A1,3 strain  2.29.1  - "axial" attack of Br + sterically and stereoelectronically favoured - group migrates trans to bromine  This result was important for a number of reasons. The reaction was fast as it went to completion in less than three hours. In addition the reaction was high yielding and gave the product as a single diastereomer. It would be informative to test this reaction on a variety of substrates. Therefore a series of substrates were synthesized and the N-bromosuccinimide promoted rearrangement reactions of these substrates were tested. 2.12 Synthesis of Substrates to be used in the N-bromosuccinimide Promoted Ring Expansion Reaction 2.12.1 Introduction Earlier in this chapter previous work conducted in the Dake lab was described. Specifically, the acid promoted semipinacol rearrangement of a number of substrates was discussed (A copy of Table 2. 1 is given below). From this work a number of observations can be made. The yields are generally good although some substrates resulted in only modest yields (51 % for entry 3). The diastereoselectivity found in the products ranges from poor (1:1.8) to very good (1:14). The reaction was very slow in some instances (144 h for entry 4) and some substrates were found to decompose under the acidic conditions. In certain cases (entries 4, 5 and 6) the products were difficult to characterize as the diastereomeric products were inseparable by chromatography. The relative configuration of these products could not be confirmed conclusively and instead were assigned by analogy with the products formed in entries 2 and 3. The N-bromosuccinimide promoted ring expansion reaction of allylic alcohol 2.1.10 to give spirocyclopentanone 2.29.1 was completely diastereoselective, high yielding and fast as it only required a few hours. This reaction was also successful where the acid conditions had caused  83  Chapter 2 A First Generation Approach Towards Halichlorine…  decomposition and the product was easily characterized. Could the N-bromosuccinimide protocol be applied to other substrates and if so, would the positive features described for the conversion of 2.1.10 to 2.29.1 apply to all substrates?  Table 2. 1 Acid Catalyzed Semipinacol Rearrangement of Allylic Cyclobutanols R1  R1  R1  R2 acid T (oC), time (h)  N Ts  R2  CH2Cl2  OH  +  N Ts  2.3.3  R2 N Ts  O 2.3.4  O 2.3.5  Entry  Substrate  R1  R2  Acida  T (°C)  t (h)  Products  % yieldb  Ratiod  1  2.3.3a  OBn  H  CSA  45  13  decomposed  N/A  N/A  2  2.3.3b  OTBS  H  CSA  45  13  2.3.4b:2.3.5b  81  2.7:1  3  2.3.3c  OPNB  H  CSA  45  13  2.3.4c:2.3.5c  51c  1:1.8  4  2.3.3d  H  Ph  CSA  45  144  2.3.4d:2.3.5d  89  1:4.5  5  2.3.3d  H  Ph  HCl  0  48  2.3.4d:2.3.5d  93  1:14  6  2.3.3.e  H  Me  HCl  0  67  2.3.4e:2.3.5e  68  1:3.7  a  CSA = (+)-camphorsulfonic acid (1.2 equiv) in dichloromethane; HCl = hydrochloric acid (1.2 equiv) in dichloromethane. b Isolated yield. c Unreacted 2.3.3c was recovered in 32 % yield. d Ratios were determined by 1H NMR integration and/or GC analysis of the product mixture. 2.12.2 Potential Substrates for the N-bromosuccinimide Promoted Ring Expansion Reaction In order to investigate the N-bromosuccinimide promoted ring expansion reaction a number of substrates were identified (Scheme 2. 30). In general substrates that contained cyclobutanols that would undergo expansion to cyclopentanones were targeted. One cyclopentanol 2.30.4 was also identified to see if the expansion of five membered rings to six membered rings could be done. The substrates fall into four categories: unsubstituted, 4-substituted, 5-substituted and 6-substituted. Compounds 2.30.1, 2.3.3e, 2.3.3d, 2.3.3a, 2.3.3b and 2.30.4 have previously been made in the Dake group. 58 The synthesis of the 4-isopropyl substrate 2.30.2 should be accessible from δ-valerolactam 2.30.3 using procedures similar to those used to make compounds 2.3.3e and 2.3.3d. The 6-allyl substrates 2.30.6 and 2.16.2 might  84  Chapter 2 A First Generation Approach Towards Halichlorine…  serve as possible intermediates in the halichlorine synthesis. These should be available from either 2.30.7 or 2.30.8.  Scheme 2. 30 Potential Ring Expansion Substrates Unsubstituted and 4-Substituted Substrates 4  Ts  N  Ts  Me  N  Ts  OH  N  Ts  2.30.1  N  OH  OH 2.3.3e  Ph  4  4  Ts  OH  2.30.2  N O  2.3.3d  2.30.3  5-Substituted Substrates OBn  5  5  Ts  OTBS  OTBS  N  Ts  5  N  OH  Ts OH  2.3.3a  OTBS  2.3.3b  N  Ts  N  OH  O  2.30.4  2.30.5  6-Substituted Substrates H N  Ts  N  OH 2.30.6  H  6  6  Ts  H  H  Ts OH  2.16.2  or  N  Ts O  2.30.7  N O 2.30.8  2.12.3 Retrosynthetic Analysis of Semipinacol Rearrangement Substrates Previous work in the Dake group has shown that allylic cyclobutanols (or cyclopentanols) such as 2.31.1 can be made by the transmetallation of alkenyl stannanes 2.31.2 followed by carbonyl addition to cyclobutanone (or cyclopentanone) (Scheme 2. 31). The alkenyl stannanes are made from the corresponding alkenyl triflates 2.31.3 which in turn come from amides 2.31.4.  85  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 31 Retrosynthetic Analysis of the Semipinacol Rearrangement Substrates  Ts  R  R  R N  Ts  N  Ts SnMe3  OH ( )n 2.31.1  2.31.2  R  N  Ts OTf  2.31.3  N O 2.31.4  2.12.4 Amide Preparation In order to begin this study a variety of amides had to be synthesized. This would require both the installation of the substituents at the various positions around the piperidine ring and the protection of the nitrogen. The p-toluenesulfonyl group was selected as the nitrogen protecting group because of its perceived compatibility with organometallic reagents, Bronsted acids and Lewis acids. While several of the amides have been previously made, their syntheses are included here for completeness. The unsubstituted amide was prepared by treating δ-valerolactam (2.31.5) with n-butyllithium and p-toluenesulfonyl chloride (Equation 2. 12).  Equation 2. 12 Synthesis of the N-(4-toluenesulfonyl)-piperidin-2-one 2.30.3 59 H  nBuLi, TsCl THF, - 78 oC to rt  N O  (82 %)  2.31.5  Ts  N O 2.30.3  The substrates having substituents at the 4-position were synthesized from 2.30.3 (Scheme 2. 32). Deprotonation of 2.30.3 with potassium hexamethyldisilazide (KHMDS) followed by addition of phenylselenenyl chloride gave selenide 2.32.1. The crude selenide was then oxidized with meta-chloroperoxybenzoic acid (m-CPBA) which resulted in an elimination reaction to provided α,β-unsaturated amide 2.32.2. The addition of a methyl, isopropyl or phenyl cuprate reagent to 2.32.2 led to the formation of the 4-substituted amides 2.32.5a, 2.32.5b and 2.32.5c respectively.  86  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 32 Synthesis of 4-Substituted Amides 2.32.3, 2.32.4 and 2.32.5  Ts  N O 2.30.3  60  R  i) KHMDS, TMSCl THF -78 oC to 0 oC ii) PhSeCl THF, -78 oC  Ts  N  SePh O  m-CPBA CH2Cl2, 0 oC (77% two steps)  Ts  2.32.1  N O 2.32.2  2 equiv RMgX CuBr SMe2 TMSCl, THF  4  Ts  N  O 2.32.3 R = Me 85 % 2.32.4 R = iPr 63 % 2.32.5 R = Ph 74 %  The synthesis of 5-substituted amide 2.30.5 was carried out by Michaël Fenster and is largely based upon work done by Herdeis and co-workers (Scheme 2. 33). 61 Diazotization of L-glutamic acid in water provided lactone 2.33.1. The carboxylic acid functional group was reduced with borane·dimethyl sulfide complex to give alcohol 2.33.2. The primary alcohol was activated towards displacement by azide ion by first converting the alcohol to a methanesulfonate group. Hydrogenation of azide 2.33.4 followed by in situ lactam formation gave piperidinone 2.33.5. The secondary alcohol was protected as its tert-butyldimethylsilyl ether and the nitrogen was protected as its p-toluenesulfonamide derivative to give the fully protected lactam 2.30.5. Scheme 2. 33 Synthesis of 5-Substituted Amide 2.30.5 H2N  H CO2H  HO 2C  L-glutamic acid  MsO  H  O  O  NaNO2, HCl o  H2O, 0 C to rt (55 %)  NaN3 18-Crown-6 CH3CN, Δ (93 %)  H N3  2.33.3  O  H HO2C  O  BH3 SMe2 THF, rt (72 %)  HO  H  2.33.1  O  2.33.4  O  H2 (3 atm) Pd/C, MeOH 30 oC (77 %)  O  2.33.2  O  MsCl, Et3N CH2Cl2 (88 %)  OH  H  N O 2.33.5  OTBS TBSCl, imidazole DMF, rt (92 %)  H  N O 2.33.6  OTBS 2.2 equiv n-BuLi 2.3 equiv TsCl THF, -78 oC (73 %)  5  Ts  N O 2.30.5  Recall from earlier in this chapter that allylation of 6-ethoxy compound 2.1.10 (Table 2. 2) and N-tosylation of 6-allyl amide 2.6.2 were not useful reactions. The synthesis of the 6-allyl substrate 2.16.2 required a new strategy. Fortunately a new route was found that would enable this compound to be made (Scheme 2. 34). An attractive feature of this new route was that it  87  Chapter 2 A First Generation Approach Towards Halichlorine…  should provide the desired 6-allyl amide in enantiopure form. Initially, only the materials to make 6-allyl lactam 2.30.7 were available and consequently the initial work involved the synthesis of lactam 2.30.7. Note that this compound has the incorrect configuration required for the synthesis of halichlorine. After acquiring the necessary materials, lactam 2.30.8 was also made. This compound has the configuration required for the halichlorine synthesis.  Scheme 2. 34 Synthesis of 6-Allyl Amides 2.30.7 and 2.30.8 HO  MeOH, H2SO4 (98 %)  O  MeO2C  O 2.34.1  i) (+) - Ipc (-) - Ipc  (COCl)2, DMSO, NEt3, CH2Cl2 -78 oC (50 %)  2.34.2  O MeO2C 2.34.3  5  2  2  or  5  PPh3, DEAD, H CO2Me , THF N Ts Boc  OH  , Et2O  ii) NaOH, H2O2, reflux  2.34.4a (5R, 59 %, >96 % ee) 2.34.4b (5S, 76 %, >96 % ee)  6  2) AlMe3  N Ts  O 2.30.7 (6S, 88 %)  N  Boc CO2Me  2.34.5a (5S, 74 %) 2.34.5b (5R, 52 %)  H 1) nitrobenzene, 180 oC  Ts  H 6  or  N Ts  O 2.30.8 (6R, 99 %)  Note: The designation of the stereogenic center changes from 5 to 6 following cyclization  The synthesis of lactams 2.30.7 and 2.30.8 started with the methanolysis of δ-valerolactone (2.34.1) to give hydroxyester 2.34.2. After oxidation of the primary alcohol to aldehyde 2.34.3 treatment with either enantiomer of B-allyldiisopinocampheylborane provided the corresponding homoallylic alcohols. 62 (The enantiomeric excess (ee) for these compounds was determined by comparing the optical rotations with those found in the literature.) Treatment with tert-butyl toluene-4-sulfonylcarbamate under Mitsunobu conditions gave homoallylic amines 2.34.5a and 2.34.5b. Thermal cleavage of the tert-butoxycarbonyl (Boc) protecting group followed by treatment with trimethylaluminum produced the desired 6-allyl lactams 2.30.7 and 2.30.8.  88  Chapter 2 A First Generation Approach Towards Halichlorine…  2.12.5 Alkenyl Triflate Formation With the desired amides in hand the lactam-derived enol triflates were then synthesized. This was done by treating the various amides with KHMDS and trapping with the appropriate triflating reagent (Table 2. 5). The yields were high (77-100 %) for all of the substrates. In some cases (2.34.6 and 2.34.7) the enol triflates converted back to the starting lactams within a few hours and so were immediately used in subsequent reactions. The 6-allyl triflates 2.34.11 and 2.34.12 could be stored in the refrigerator as solutions in diethyl ether containing ~1 % triethylamine.  Table 2. 5 Formation of Alkenyl Triflates R1 R 2  R3 R4  R 1 R2  R4  KHMDS, ArNTf 2 THF, -78 oC to rt  N  R3  Ts  O Substrate  Entry Substrate  N Ts  OTf Product  R1  R2  R3  R4  Product Yield (%)  1  2.30.3  H  H  H  H  2.34.6  87  2  2.32.3  H  H  H  Me  2.34.7  80  3  2.32.4  H  H  H  iPr  2.34.8  81  4  2.32.5  H  H  H  Ph  2.34.9  77  5  2.30.5  H  H  OTBS  H  2.34.10  76  6  2.30.7  H  Allyl  H  H  2.34.11  95  7  2.30.8  Allyl  H  H  H  2.34.12  100  2.12.6 Alkenyl Stannane Formation The conversion of the enol triflates to the corresponding alkenyl stannanes was typically done using hexamethyldistannane in the presence of a catalytic amount of tris(bisdibenzylideneacetone)dipalladium (0) (Pd2dba3) with triphenylarsine in THF at rt for 7-9 h (Table 2. 6, entries 1-6). 63 Longer reaction times appeared to lead to decomposition of the alkenyl stannane product. The yields for this conversion ranged from 24 to 71 %. Particularly noteworthy is that enol triflate 2.34.10 gave only a 24 % yield of the desired alkenyl stannane  89  Chapter 2 A First Generation Approach Towards Halichlorine…  2.34.18. Because alkenyl stannane was intended to be used in a possible synthetic route towards halichlorine an alternative procedure was sought to improve this yield.  Table 2. 6 Synthesis of Alkenyl Stannanes R1 R 2  (Me3Sn)2 n=5 or 10 mol % Pd2dba3  R3 R4  R1 R2  OTf Substrate  Entry Substrate R1  R4  N  4n mol% AsPh3 THF, rt  N Ts  R3  Ts  SnMe3 Product  R2  R3  R4  Product Yield (%)  1  2.34.6  H  H  H  H  2.34.13  65  2  2.34.7  H  H  H  Me  2.34.14  42  3  2.34.8  H  H  H  iPr  2.34.15  41  4  2.34.9  H  H  H  Ph  2.34.16  45  5  2.34.10  H  H  OTBS  H  2.34.17  71  6  2.34.11  H  Allyl  H  H  2.34.18  24  2.12.7 Alternative Alkenyl Stannane Procedure As mentioned above the palladium catalyzed coupling reaction between the lactam derived enol triflate 2.34.11 and hexamethyldistannane was quite low yielding. Improving this yield was important because this reaction was to be used in the total synthesis of halichlorine. Therefore, an alternative set of conditions was investigated.  Scheme 2. 35 Coupling of Enol Triflates with Gilman Reagents OTf  R R2CuLi, THF  R = Me, Bu, Ph, vinyl, cyclopropyl  - 15 oC (62 - 100 %) 2.35.1 TfO  CuL2 R  sigma bond metathesis  2.35.2  Chapter 2 A First Generation Approach Towards Halichlorine…  90  In 1980, McMurry showed that lithium dialkylcuprates (Gilman reagents) could be coupled with ketone-derived enol triflates (Scheme 2. 35). 64 A number of plausible mechanisms have been proposed for this type of transformation and for related processes. These include: a) a process that involves π-complex intermediates 65, b) a free radical mechanism, c) a single electron transfer mechanism 66, d) a process involving oxidative addition and reductive elimination 67 and e) a sigma bond metathesis mechanism 68. Venkataraman and van Allen have done a mechanistic study of the Ullman reaction, which is a copper mediated coupling reaction between an aryl halide and a phenol or an amine.68 Their studies, combined with the work of others, seem to indicate that a sigma bond metathesis mechanism is the most likely mechanism for this transformation. It is worth noting that the reactions conducted by McMurray are not Ullmann coupling reactions however it is reasonable to believe that both reactions proceed through similar mechanisms. The products formed in McMurray’s studies involve the overall insertion of an alkyl group from a cuprate reagent into a carbon triflate bond. To compare this with our desired transformation it would be necessary to a) make a trimethylstannyl cuprate reagent and b) find out if the trimethylstannyl group can insert into the carbon triflate bond. Since McMurray’s initial work there have been a limited number of copper-mediated coupling reactions that formally insert a trialkyltin group into a carbon-triflate bond. 69 Our studies focused on using lithium trimethylstannyl copper (I) cyanide to do the desired transformation. This reagent has been widely used in the Piers group to make highly functionalized alkenes from alkynoates, although they reported only one example where the trimethyltin moiety inserted into a carbon-triflate bond. 70 It should be noted that in the one example that was attempted, the substrate contained a triflate on the β-carbon of an α,βunsaturated ester and therefore the mechanism might be different than that for the reaction of an isolated triflate. Fortunately, this reaction worked quite well for the desired transformation (Table 2. 7). Hence treatment of enol triflate 2.34.11 with three equivalents of lithium trimethylstannyl copper (I) cyanide in THF at 0 °C for 5 hours resulted in the formation of alkenyl stannane 2.34.18 in 73 % yield (entry 2). When two equivalents of the cuprate reagent were used only a 51 % yield of the desired alkenyl stannane was obtained (entry 1). The reaction worked equally well with enol triflate 2.34.12, i.e. the enantiomer of 2.34.11 (entry 3).  Chapter 2 A First Generation Approach Towards Halichlorine…  91  Table 2. 7 Formation of Alkenyl Stannanes Using A Cuprate Reagent R1  R2  R1  X equiv (SnMe3CuCN)Li THF, 0 oC, 5 h  N Ts  N Ts  OTf  Substrate  Entry Substrate  R2  SnMe3 Product  R1  R2  X Product Yield (%)  1  2.34.11  H  allyl  2  2.34.18  51  2  2.34.11  H  allyl  3  2.34.18  73  3  2.34.12  allyl  H  3  2.34.19  65  Alkenyl stannanes are used widely in organic chemistry, most notably as one of the coupling partners in Stille cross coupling reactions. In certain cases, using palladium catalysis to convert enol triflates to alkenyl stannanes either does not work or is low yielding. The procedure described above offers an alternative method for the synthesis of alkenyl stannanes. This procedure also solved an important problem in our synthetic route towards halichlorine. 2.12.8 Synthesis of the Allylic Cyclobutanols  Most of the allylic alcohols were synthesized using the standard protocol developed in our laboratories (Table 2. 8). The transmetallation of the alkenyl stannane to the alkenyllithium species generally required at least 2 equivalents of methyllithium in diethyl ether at -78 °C. Addition of anhydrous magnesium bromide·diethyl etherate prior to addition of cyclobutanone was required in most cases to obtain satisfactory yields of the allylic cyclobutanols. The yields for this reaction varied from 74 to 89 %.  92  Chapter 2 A First Generation Approach Towards Halichlorine…  Table 2. 8 Formation of Allylic Cyclobutanols R1 R2  R3 R4  2  N Ts  R1 R 2  i) MeLi, Et2O, -78 oC to 0 oC ii) MgBr .OEt , Et O, -78 oC iii)  SnMe3  2  2  Ts  ( )n  Substrate  Entry Substrate  R4  N  , Et2O, -100 oC to rt  O  R3  OH ( )n Product  R1  R2  R3  R4  n Product Yield (%)  H  1  2.30.1  89  1  2.34.13  H  H  H  2  2.34.14  H  H  H  Me 1  2.3.3e  85  3  2.34.15  H  H  H  iPr  1  2.30.2  83  4  2.34.16  H  H  H  Ph  1  2.3.3d  85  5  2.34.17  H  H  OTBS  H  1  2.3.3b  74  5  2.34.17  H  H  OTBS  H  2  2.30.4  84  6  2.34.18  H  allyl  H  H  1  2.30.6  67  7  2.34.19  allyl  H  H  H  1  2.16.2  61  The 5-benzyloxy substrate was synthesized from 5-tert-butyldimethylsilyloxy substrate 2.30.7 (Scheme 2. 36). The silyl ether was cleaved by treatment with tetrabutylammonium fluoride in THF at rt. The resulting secondary alcohol was then converted to the benzyl ether by deprotonation with sodium hydride and quenching with benzyl bromide in the presence of tetrabutylammonium iodide in THF solution.  Scheme 2. 36 Synthesis of 5-Benzyloxy Substrate 2.3.3a OBn  OTBS 1) TBAF, THF, rt 2) NaH, BnBr, TBAI, THF (86 % two steps)  N Ts  OH  2.3.3b  N Ts  OH 2.3.3a  2.12.9 Summary A number of ring expansion substrates have been identified and their syntheses have been described. Specifically, ring expansion substrates have been made that contain different substituents at the 4-, 5-, and 6-positions of the piperidine ring. Additionally, a new method for  Chapter 2 A First Generation Approach Towards Halichlorine…  93  the conversion of enol triflates to the corresponding alkenyl stannanes has been discovered. This transformation uses a stoichiometric amount of lithium trimethylstannyl copper (I) cyanide to provide good yields of alkenyl stannanes. 2.13 N-Bromosuccinimide Promoted Ring Expansion Reactions With the necessary substrates in hand, the N-bromosuccinimide promoted ring expansion reactions were examined (Table 2. 9). The reaction was successful in all cases giving high yields of the 6-azaspirocycles (79-100%). The general structures of the products were confirmed by several pieces of diagnostic evidence. The presence of the ketone was supported by both infrared (IR) and 13 C NMR data. Specifically, the IR spectra showed strong signals in the range 1740-1759 cm-1 while the 13 C NMR spectra gave signals in the range of δ 208.2-216.2. In cases where mass spectral analysis was done there were always two parent mass peaks observed: one containing the 79Br isotope and one containing the 81Br isotope. 71 The diastereoselectivities of the ring expansion reactions ranged from little selectivity (entries 2, 3 and 5) to complete selectivity (entries 1, 4, 7, and 8). Increasing the size of the substituent at the 4-position from methyl to isopropyl to phenyl resulted in an increase in diastereoselectivity (entries 2,3 and 4). 72 In each case the major diastereomer was the one where the bromine is trans to the substituent at the 4-position. Substrates with oxygen substituents at the 5-position reacted much less selectively overall. The 5-benzyloxy derivative 2.3.3a produced a 5:1 mixture of 2.36.5a and 2.36.5b while the 5-tert-butyldimethylsilyloxy substituted compound 2.3.3b provided 2.36.6a and 2.36.6b in a 3.5:1 ratio. In both cases the ether substituent in the major product was trans to the bromine substituent. Similar to when 2.1.10 was treated with N-bromosuccinimide the reactions of the 6-allyl substrates resulted in the formation of a single diastereomeric product. The major product was the one in which the bromine is trans to the allyl group. A complete discussion regarding stereochemical assignments will be given below.  94  Chapter 2 A First Generation Approach Towards Halichlorine…  Table 2. 9 Ring Expansions Promoted by N-Bromosuccinimide R1 R2  R3  O  R4  N  i  Ts  Br N  OH  R1 R2  O  R1 R 2  R4  N  O  PrOH,  R3  Br O  Ts  +  a  R3 R4  N  Br O  Ts  -78 οC to rt Substrate  a  b  Yield  Ratio  (%)b  a:bc  2.36.1a  85  1:0  Me  2.36.2a/b  79  1.9:1  H  iPr  2.36.3a/b  85  3.5:1  H  H  Ph  2.36.4a  80  1:0  H  H  OBn  H  2.36.5a/b  90  5:1  2.3.3b  H  H  OTBS  H  2.36.6a/b  95  3.5:1  7  2.30.6  H  allyl  H  H  2.36.7b  100  1:0  8  2.16.2  allyl  H  H  H  2.36.8a  100  1:0  Entry  Substrate  R1  R2  R3  R4  Product(s)  1  2.30.1  H  H  H  H  2  2.3.3e  H  H  H  3  2.30.2  H  H  4  2.3.3d  H  5  2.3.3a  6  a  Standard conditions: NBS (1.2 equiv) in a 1:1 mixture of 2-propanol and propylene oxide at -78 °C. b Isolated yields. c Ratio was determined by the 1H NMR spectrum of the crude reaction mixture and typically confirmed by isolation of each of the diastereomeric products. Unfortunately the attempted ring expansion of cyclopentanol 2.30.4 was not successful (Equation 2. 13). Treatment of this compound using the standard conditions described in Table 2.9 provided alkenyl bromide 2.36.9 in 75 % yield.  Equation 2. 13 Ring Expansion of Cyclopentanol 2.30.4 Fails OTBS  Ts  N  PrOH,  i  OH  2.30.4  O  Br N  OTBS  O O  Ts  N  ο  -78 C to rt 2.36.9  Br OH  95  Chapter 2 A First Generation Approach Towards Halichlorine…  2.13.1 Stereochemical Assignments 2.13.1.1 Introduction The assignment of the configuration of the products was done by using 1H NMR spectroscopic analysis, single crystal X-ray analysis or by analogy with structures whose configuration had already been determined. 2.13.1.2 Substituent at the 4-Position As mentioned above, the 4-phenyl substituted compound 2.3.3d reacted to form 2.36.4a as a single diastereomer. The configuration of this compound was ultimately confirmed by single crystal X-ray analysis (Scheme 2. 37). The ORTEP diagram clearly shows that the piperidine ring sits in a chair conformation where the bromine and the ketone are on the same face of the 6-membered ring. This is consistent with migration occurring antiperiplanar to the bromine. The bromine and the phenyl substituents end up trans to one another and reside in equatorial orientations. Interestingly, the p-toluenesulfonyl group resides in an axial orientation. Scheme 2. 37 Ring expansion and ORTEP of Compound 2.36.4a Ph  Ph Ts  NBS  N  N OH  2.30.4  Ts  Br O  OO S N  H Br H O  2.36.4a  The relationship between the bromine and the phenyl group is supported by NMR data. The coupling constant (J=11.6 Hz) between the protons attached to C10 and C9 of compound 2.36.4a is consistent with a transdiaxial relationship (Table 2. 10). Cyclopentanones 2.36.2a (methyl substituent) and 2.36.3a (isopropyl substituent), which were the major diastereomers formed in their respective reactions, also had similar coupling constants (J=11.0 Hz for 2.36.2a and J=10.6 Hz for 2.36.3a). The coupling constants between the corresponding protons in the minor diastereomers were much smaller (no coupling observed for 2.36.2b and J=1.8 Hz for 2.36.3b) (Table 2. 11). The C=O IR stretching frequencies and the chemical shifts for the  96  Chapter 2 A First Generation Approach Towards Halichlorine… 13  carbonyl carbons in the C NMR spectra are also quite informative. The major products have C=O IR stretching frequencies ranging from 1740-1744 cm-1 while the minor products have much higher stretching frequencies ranging from 1755-1759 cm-1. In the 13C NMR spectra the chemical shifts for the carbonyl carbons of the major products range from δ 213.1-214.0 ppm while in the minor products the chemical shifts are lower ranging from δ 208.2-208.3 ppm.  Table 2. 10 Spectroscopic Data Used to Establish the Structures of 2.36.2a, 2.36.3a and 2.36.4a R  Ts H N  Br O  Br * O  10  N Ts  9  R  H  δH10 (ppm) m J(Hz) C=O IR stretch (cm-1)  δC* (ppm)  Major Product  R  2.36.2a  Me  3.99  d  11.0  1743  213.6  2.36.3a  iPr  4.16  d  10.6  1744  214.0  2.36.4a  Ph  4.52  d  11.6  1740  213.1  Table 2. 11 Spectroscopic Data Used to Establish the Structures of 2.36.2b, 2.36.3b and 2.36.4b R N Ts  Ts Br N  O *  Br O  R  2.36.2b  Me  4.01  d  2.36.3b  iPr  4.15  d  2.36.4b  Ph  10  9  R  H  C=O IR stretch (cm-1)  δC* (ppm)  n/a  1755  208.2  1.8  1759  208.3  δH10 (ppm) m J(Hz)  Minor Product  H  n/a  It would be useful to determine the relative configuration of one of the minor products resulting from ring expansion of one of the 4-substituted compounds. This would allow us to confirm or refute the hypothesis that the minor diastereomer formed in these reactions is the one resulting from electrophilic attack of bromine onto the opposite face of the alkene followed by migration of the alkyl group antiperiplanar to the bromonium ion. In other words we wanted to rule out that ring expansion products resulting from syn migration had occurred. Unfortunately,  Chapter 2 A First Generation Approach Towards Halichlorine…  97  none of the minor products were crystalline and therefore single crystal X-ray analysis was not possible. Fortunately, reduction of 2.36.2b to alcohol 2.37.1 gave a solid compound that after recrystallization gave X-ray quality crystals (Equation 2. 14). Single crystal X-ray analysis confirmed that our structural assignment was correct. The bromine, the methyl group and the alcohol are all cis with respect to one another. Based upon this information and the results from above the configuration of the 4-isopropyl products 2.36.3a and 2.36.3b was assigned by analogy.  Equation 2. 14 Reduction and Confirmation of the Relative Configuration of 2.36.2b O  Ts Br N H  Me  LiEt3BH, THF, rt (70 %)  HO  Ts Br N H  H  2.36.2b  Me H  2.37.1  2.13.1.3 Substituent at the 5-position The spirocyclic carbon centers for compounds 2.36.6a and 2.36.6b were assigned (ketone carbonyl cis to the bromine) by analogy with the 4-substituted products. The relative configuration of these compounds was established by examining the coupling constants of specific protons in conjunction with nuclear Overhauser enhancement (NOE) studies (Figure 2. 12). For each of the compounds 2.36.6a and 2.36.6b, the protons on the carbon bearing the bromine substituent (Ha and Hc) were assigned to be pseudoaxial on the basis of a large coupling constant for each (Ha: δ 4.21 ppm, dd, J=9.9 and 4.0 Hz; Hc: δ 4.21 ppm, dd, J=12.8, 4.6 Hz). The proton on the carbon bearing the silyl ether in either compound 2.36.6a or 2.36.6b (Hb or Hd) could not be as trivially assigned as being axial or equatorial. A significant NOE was present between Hc and Hd in 2.36.6b while no NOE was observed between Ha and Hb in 2.36.6a.  98  Chapter 2 A First Generation Approach Towards Halichlorine…  Figure 2. 12 NMR Data Used to Establish the Structures of 2.36.6a and 2.36.6b a  O Ts  OTBS  Ts H N  b  H  Br O  N  Br  a  c  δ H 4.21 ppm J=12.8, 4.6 Hz  δ H 4.21 ppm J=9.9, 4.0 Hz  H  c  OTBS H  d  strong NOE 2.36.6b  2.36.6a  Unfortunately, the products resulting from the ring expansion of 2.30.6, compounds 2.36.5a and 2.36.5b, were inseparable by chromatographic methods. Initially attempts were made to convert the benzyl ethers of compounds 2.36.5a and 2.36.5b into the corresponding silyl ethers. If successful then the resulting 1H NMR spectra of these compounds could be correlated with those obtained for compounds 2.36.6a and 2.36.6b. Unfortunately the benzyl protecting groups proved to be difficult to remove and therefore the correlation could not be made in this manner. Next an attempt was made to convert the silyl ether of compound 2.36.6a into the corresponding benzyl ether, again with the intention of correlating the 1H NMR spectrum with that obtained for the mixture of compounds 2.36.5a and 2.36.5b. Deprotection of the silyl ether of 2.36.6a was easily done by stirring the compound in acidic ethanol (Equation 2. 15). Equation 2. 15 Deprotection of Silyl Ether 2.36.6a OH  OTBS  Ts  N  Br O 2.36.6a  1% HCl in EtOH (quantitative)  Ts  N  Br O  2.37.2  Protection of the secondary alcohol as its benzyl ether proved to be surprisingly difficult (Table 2. 12). All attempts to benzylate the secondary alcohol using a base and benzyl bromide either resulted in no reaction or led to decomposition of the starting material (entries 1-5). Treatment of the alcohol with benzaldehyde in the presence of a Lewis acid and triethylsilane also failed (entry 6). In the preceding reaction involving the deprotection of the silyl ether of 2.36.6a it was found that alcohol 2.37.2 was stable to acidic conditions. Consequently experiments were conducted that involved introduction of the benzyl ether under acidic  99  Chapter 2 A First Generation Approach Towards Halichlorine…  conditions. Treatment of the alcohol with benzyltrichloroacetimidate in the presence of several Lewis acids either resulted in no reaction or led to decomposition of the starting material (entries 7-10). Fortunately when the alcohol was treated with phenyldiazomethane and tetrafluoroboric acid a modest yield (44 %) of the desired benzyl ether was obtained (entry 11). The 1H NMR spectrum of this compound was identical to that obtained for 2.36.5a. In this manner it was proved that the major diastereomer from the ring expansion of 2.3.3a was the same as the major diastereomer formed in the ring expansion of 2.3.3b.  Table 2. 12 Conversion of 2.37.2 to 2.36.5a OBn  OH Conditions Ts  N  Br O 2.37.2  Entry  Ts  N  Br O  2.36.5a  Conditions i) NaH, THF, 0 °C 1 ii) BnBr, TBAI, 0 °C – rt NaH, DMF, 0 °C 2 BnBr, TBAI, 90 °C KHMDS, THF 3 BnBr, TBAI, reflux KH, THF, 0 °C 4 BnBr, TBAI, 0 °C – rt i) KH, THF, 0 °C 5 ii) BnBr, TBAI, 18-c-6, 0 °C - rt TMSOTMS, TMSOTf, 6 benzaldehyde, Et3SiH benzyltrichloroacetimidate 7 F3CSO3H, CH2Cl2, cyclohexane benzyltrichloroacetimidate 8 F3CSO3H, CH2Cl2 benzyltrichloroacetimidate 9 CSA, CH2Cl2 benzyltrichloroacetimidate 10 PPTs, CH2Cl2 BnN2, HBF4, 11 CH2Cl2, -40 °C a The starting material was recovered.  2.36.5a (%) 0a decomposition decomposition decomposition decomposition 0a decomposition decomposition 0a 0a 44  Chapter 2 A First Generation Approach Towards Halichlorine…  100  2.13.1.4 Substituent at the 6-position The configuration of the products resulting from the ring expansion of the 6-allyl substrates was either confirmed by single crystal X-ray analysis or through NMR data. The configuration of 6-allyl compound 2.36.7 was confirmed by single crystal X-ray analysis (Scheme 2. 38). The ORTEP diagram shows that the piperidine ring sits in a twist boat conformation where the bromine and the ketone are cis. Again this is consistent with antiperiplanar attack on the bromine π-complex or the bromonium ion. The bromine resides in a pseudoequatorial orientation while the 6-allyl substituent occupies a pseudoaxial position. It is noteworthy that the 6-allyl substituent and the ketone are on opposite faces of the piperidine ring as this would have implications for our total synthesis of halichlorine. This will be discussed later in the chapter. The ring expansion of the enantiomer of 2.30.6, compound 2.16.2 gives the enantiomeric product of 2.36.7b, i.e. compound 2.36.8a. Scheme 2. 38 Confirmation of Configuration for 2.36.7b  Ts  O  NBS  N  N OH  2.30.10  Ts  O Br O  S  N Br O H  H  2.36.7b  2.13.2 Rationalization of Diastereoselectivity 2.13.2.1 Substituent at the 4-Position Presumably the most stable ground state half-chair conformation of the starting material for the 4-substituted substrates would be one where the R group occupies a pseudoequatorial orientation (Scheme 2. 39). If the reaction proceeds through the most stable ground state halfchair conformation then for steric and stereoelectronic reasons the electrophilic bromine should approach the alkene on the same face of the piperidine ring as the R group. Subsequent rearrangement leads to the formation of the cis,cis product as the major diastereomer. 73 If the reaction is occurring through a half-chair conformation then the half-chair conformation where  101  Chapter 2 A First Generation Approach Towards Halichlorine…  the R group is pseudoaxial must be preferred over the one in where the R group is pseudoequatorial. This leads to the formation of the trans,cis diastereomer.  Scheme 2. 39 Proposed Mechanism for the Ring Expansion of 4-Substituted Substrates Favoured R Ts N  R H N  OH  Ts  Br O  +  R Ts  N  Br - R group pseudoaxial - presumably a less stable ground state conformation - more reactive transition state  trans,cis Major Diastereomer  OH +  Br  R  OH  N R  Ts N  Ts  Br O  H - R group pseudoequatorial - presumably the most stable ground state conformation - less reactive transition state  cis,cis Minor Diastereomer  An alternative way of thinking about this reaction is that perhaps the electrophilic bromine approaches the less hindered face of the alkene, i.e. the face opposite to the R group. 74 Pseudoaxial attack of the bromine would place the R group in a pseudoaxial orientation. The ring expansion occurs wherein the migrating alkyl group is oriented trans to the bromine atom on the piperidine ring. 2.13.2.2 Substituent at the 5-Position For the 5-substituted substrates, the major product was the one in which the bromine was trans to the 5-substituent. Presumably the most stable ground state half-chair conformation should be the one where the substituent occupies a pseudoequatorial conformation (Scheme 2. 40). “Pseudoaxial approach” of NBS to these substrates followed by antiperiplanar migration of the alkyl group leads to the observed major diastereomer. However, the selectivity in these  102  Chapter 2 A First Generation Approach Towards Halichlorine…  reactions was not great therefore it must be assumed that there is only a slight stereoelectronic preference for the reaction to occur via the most stable ground state half-chair conformation.  Scheme 2. 40 Proposed Mechanism for the Ring Expansion of the 5-Substituted Substrates Favoured OR Ts N  OR H  OH  N  Br O  Ts  +  Br  OR  Ts  Major Diastereomer  N OH Br OH  OR  +  OR  Ts N  N H  Ts  Br O  Minor Diastereomer  2.13.2.3 Substitution at the 6-Position The diastereoselectivity of the NBS promoted ring expansion reaction was complete for all substrates with substitution at the 6-position. The explanation provided for the conversion of 2.1.10 to 2.29.1 can be invoked to explain the results for all of the 6-substituted substrates. The substituent at the 6-position probably resides in a pseudoaxial orientation so that A1,3 strain with the p-toluenesulfonyl group on the nitrogen is minimized (Scheme 2. 41). N-Bromosuccinimide approaches the alkene on the sterically less hindered face of the more stable half-chair conformation. Formation of a bromonium ion or a bromine π-complex triggers an alkyl shift, which after loss of a proton results in the formation of the cyclopentanone. The results are consistent with antiperiplanar migration of the alkyl group where the migrating group and the bromine end up on opposite faces of the piperidine ring.  103  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 41 Proposed Mechanism for the Ring Expansion of the 6-Substituted Substrates Favoured  R R  H Ts N  Ts N  OH Br  Br O  +  Only Isolated Product R Ts  N OH Br OH Ts RN  +  R  x  Ts N  Br O  H  2.13.3 Summary Up until now the use of electrophilic halogen reagents to promote ring expansion reactions had not been widely studied. The N-bromosuccinimide promoted ring expansion reaction has been shown to be a powerful method for the construction of 6-azaspirocyclic ketones. This protocol has been applied to a variety of substrates giving 6-azaspirocyclic ketones in high yields. The diastereoselectivity in the products ranges from poor to excellent. There is substantial evidence to support the fact that the alkyl group migrates anti to the bromine atom. The products formed in this reaction are generally easy to purify by chromatographic methods and characterization of the products seems to be easier than was found for the products produced in the acid catalyzed semipinacol rearrangement. 2.14 Additional Substrates for the Acid Catalyzed Semipinacol Rearrangement and the Verification of the Configuration of the Products 2.14.1 Introduction Earlier in this chapter the acid catalyzed semipinacol rearrangement reactions of several substrates were described (Table 2. 1). From the work involving N-bromosuccinimide more substrates were now available. Specifically the acid catalyzed rearrangement reactions with  104  Chapter 2 A First Generation Approach Towards Halichlorine…  compounds 2.30.2 and 2.30.6 was attempted. Because the relative configuration of the products formed in the N-bromosuccinimide promoted ring expansion reactions had been determined it was hoped that some of these products could be used to help verify the relative configuration of the products formed in the acid catalyzed ring expansion reactions.  Table 2. 1 Acid Catalyzed Semipinacol Rearrangement of Allylic Cyclobutanols R1  R1  R1  R2 acid T (oC), time (h)  N Ts  R2  CH2Cl2  OH  +  N Ts  2.3.3  R2 N Ts  O 2.3.4  O 2.3.5  Entry  Substrate  R1  R2  Acida  T (°C)  t (h)  Products  % yieldb  Ratiod  1  2.3.3a  OBn  H  CSA  45  13  decomposed  N/A  N/A  2  2.3.3b  OTBS  H  CSA  45  13  2.3.4b:2.3.5b  81  2.7:1  3  2.3.3c  OPNB  H  CSA  45  13  2.3.4c:2.3.5c  51c  1:1.8  4  2.3.3d  H  Ph  CSA  45  144  2.3.4d:2.3.5d  89  1:4.5  5  2.3.3d  H  Ph  HCl  0  48  2.3.4d:2.3.5d  93  1:14  6  2.3.3.e  H  Me  HCl  0  67  2.3.4e:2.3.5e  68  1:3.7  a  CSA = (+)-camphorsulfonic acid (1.2 equiv) in dichloromethane; HCl = hydrochloric acid (1.2 equiv) in dichloromethane. b Isolated yield. c Unreacted 2.3.3c was recovered in 32 % yield. d Ratios were determined by 1H NMR integration and/or GC analysis of the product mixture. 2.14.2 Acid Catalyzed Sempinacol Rearrangements The acid catalyzed semipinacol rearrangement reactions of allylic alcohols 2.30.2 and 2.30.6 was successful in each case (Table 2. 13). The reactions were quite slow as the reactions required in excess of two days to go to completion. The acid catalyzed rearrangement of allylic alcohol 2.30.2 resulted in the formation of an inseparable mixture of two diastereomeric ring expansion products 2.41.1a and 2.41.1b. These were formed in a ratio of 4.7:1 favouring spirocyclopentanone 2.41.1a. The acid catalyzed rearrangement of allylic alcohol 2.30.6 gave only one diastereomeric ring expansion product 2.41.2b however a number of undetermined byproducts were also formed during the reaction. Consequently the yield of 2.41.2b was somewhat low (47 %).  105  Chapter 2 A First Generation Approach Towards Halichlorine…  Table 2. 13 Acid Catalyzed Semipinacol Rearrangement of 2.30.2 and 2.30.6 R2  R1 N Ts  OH  R2  R1 HCl rt, time (h) CH2Cl2  +  N Ts  N Ts  O  Substrate  R2  R1  a  O b  Entry  Substrate  R1  R2  t (h)  Products  % yieldb  Ratiod  1  2.30.2  H  iPr  50  2.41.1a,2.41.1b  71  1:4.7  2  2.30.6  allyl  H  55  2.41.2b  47  1  2.14.3 Stereochemical Assignments The relative configurations of 2.41.1a and 2.41.1b could not be determined conclusively as the products were inseparable. The relative configuration of 2.41.2b was also difficult to determine. This was not surprising considering that the configurations of some of the other products formed in the acid catalyzed semipinacol rearrangement could not be determined. Perhaps the products formed in the acid catalyzed ring expansion reactions could be chemically correlated with the products formed in the N-bromosuccinimide promoted ring expansion reactions. If so then the relative configuration of the products formed in the acid catalyzed semipinacol rearrangement reactions could be determined conclusively. 2.14.3.1 Substituent at the 4-Position In the acid catalyzed semipinacol rearrangement of substrates with groups at the 4-position the configuration of the products could not be determined conclusively. This was because the diastereomeric products were inseparable. It was assumed that the major diastereomers were the ones in which the R groups were cis to the ketone (Scheme 2. 42). This assumption would be correct if the reaction proceeds through the most stable half-chair conformation in which the R group is equatorial. However, the results from the N-bromosuccinimide promoted ring expansion reactions involving substrates with groups at the 4-position indicate that the major products are those in which the R group and the ketone are trans. Based upon this we wondered if the correct stereochemical assignments were made for the products formed in the acid catalyzed semipinacol rearrangement reactions of substrates with groups at the 4-position.  106  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 42 Assumed Product Outcome for the Acid Catalyzed Rearrangement of Substrates with Substituents at the 4-position R Ts N  R H  - Ketone and R group trans - Assumed minor diastereomer  N  OH  Ts  O  +  H R Ts  N OH +  H  R  OH  N R  Ts N  Ts  O  - Ketone and R group cis - Assumed major diastereomer  H Most stable half-chair conformation  The products resulting from the N-bromosuccinimide promoted reactions were separable by column chromatography and consequently could be isolated as single diastereomers. An attempt was made to convert the compounds formed in the N-bromosuccinimide promoted ring expansion reactions into the products formed in the acid catalyzed rearrangement reactions. This would confirm or refute the original stereochemical assignments. One way to correlate the compounds formed in the N-bromosuccinimide promoted ring expansion reactions with those formed in the acid catalyzed ring expansion reactions would be to take the products formed in the N-bromosuccinimide promoted ring expansion reactions and exchange the bromine atoms for hydrogen atoms. This was done with compounds 2.36.2a, 2.36.2b and 2.36.4a (Scheme 2. 43). Ketones 2.36.2a, 2.36.2b and 2.36.4a were initially reduced with lithium triethylborohydride (Superhydride™). Treatment of the alcohols with trin-butyltin hydride and AIBN in refluxing benzene gave the desired debrominated compounds. 75 Subsequent oxidation with tetrapropylammonium peruthenate (TPAP) and N-methylmorpholine-N-oxide restored the ketone functionality. When the 1H NMR spectra of these compounds were compared with those obtained from the acid catalyzed semipinacol reactions it was found that the major products formed in the N-bromosuccinimide promoted ring expansion reations, compounds 2.36.2a and 2.36.4a, were correlated with the minor products formed in the acid catalyzed semipinacol reactions, compounds 2.3.4e and 2.3.4d, respectively. Conversely, compound 2.36.2b, the minor product formed in the N-bromosuccinimide promoted ring expansion reation of allylic cyclobutanol  107  Chapter 2 A First Generation Approach Towards Halichlorine…  2.3.3e, was correlated with compound 2.3.5e, the major product formed in the acid catalyzed ring expansion reaction. Therefore, the original stereochemical assignments made for the compounds formed in the acid catalyzed ring expansion reactions with substituents at the 4-position were correct. The configurations of the 4-isopropyl products were assigned by analogy with the results obtained for the 4-methyl and 4-phenyl compounds.  Scheme 2. 43 Stereochemical Assignments for Compounds 2.3.4d, 2.3.4e and 2.3.5e R N  Ts H N  Br O  Ts  Br O  R  R  TPAP, NMO  2.43.2a (R = Me, 99 %) 2.43.2b (R = Ph, 76 %) O N  Ts H N  LiEt3BH (70 %)  N Ts  HO  O  Ts Br N H  H  2.36.2b HO  PhH, Δ  R  R H H O 2.3.4e (R = Me, 76 %) 2.3.4d (R = Ph, 67 %)  Ts Br N H  Br O  AIBN, nBu3SnH,  2.43.1a (R = Me, 90 %) 2.43.1b (R = Ph, 80 %)  H H OH  Ts  R  Br H OH  H  2.36.2a (R = Me) 2.36.4a (R = Ph) Ts H N  Ts H N  LiEt3BH  AIBN, nBu3SnH, PhH, Δ (71 %)  H  2.43.1c  Ts H N H  2.43.2c  O TPAP, NMO (77%) H  Ts H N H  N H  Ts  O  2.3.5e  Overall, this means that the major products formed in the acid catalyzed ring expansion reactions have the opposite relative configuration compared with the major products formed in the N-bromosuccinimide promoted ring expansion reactions. Therefore these methods are complimentary because they provide access to both diastereomers of 6-azaspirocyclic ketones with substituents at the 4-position. The debromination sequence outlined in Scheme 2. 43 provided the first concrete evidence for the stereochemical assignments made for the products resulting from the acid catalyzed semipinacol rearrangement of substrates with substituents at the 4-position.  108  Chapter 2 A First Generation Approach Towards Halichlorine…  2.14.3.2 The 6-Allyl Product 2.41.2b The product formed from the N-bromosuccinimide promoted ring expansion reaction of 2.30.6 was the one in which both the ketone and the bromine substituents were trans to the allyl substituent. The product formed in the acid catalyzed semipinacol rearrangement was expected to give the same relative configuration, i.e. where the ketone is trans to the allyl group. For the reasons mentioned earlier, the allyl group should be oriented pseudoaxial in the lowest energy half-chair conformation (Scheme 2. 44). This helps to minimize A1,3 strain with the adjacent N-tosyl group. Protonation of the double bond forms an intermediate resonance-stabilized azacarbenium ion. The migrating group should approach the azacarbenium ion from a pseudoaxial orientation. This would place the migrating group cis to the 6-allyl substituent and consequently the ketone would end up trans to the 6-allyl group. Chemical correlation was used to confirm this. Scheme 2. 44 Model Used to Predict the Configuration of the Product Formed in the Acid Catalyzed Semipinacol Rearrangement of Allylic Alcohol 2.30.6 intermediate aza-carbenium ion H+ Ts  N  HO  N +  Ts H  OH  Ts N  - allyl group and ketone are trans O  - allyl group axial - group migrates to pseudoaxial position  When bromide 2.36.7b was treated with 1,8-diazabicyloc[5.4.0]undec-7-ene (DBU) in refluxing benzene, the E2 elimination product 2.45.1 was obtained (Scheme 2. 45). Subsequent hydrogenation resulted in the formation of compound 2.45.2. When the acid catalyzed ring expansion product 2.41.2a was hydrogenated compound 2.45.2 was also formed. This outcome was supported by the fact that the 1H NMR spectra for the two hydrogenation products were identical. Therefore the products formed in both the N-bromosuccinimide promoted ring expansion reaction and the acid promoted ring expansion reaction of the 6-allyl substrate have the same relative configuration.  109  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 45 Chemical Correlation of 2.36.7b and 2.41.2b H  H  H  Ts N  Br O  DBU, PhCH3 reflux (27 %)  2.36.7b H Ts N O  H2, Pd/C MeOH (89%)  2.41.2b  Ts N O  H2, Pd/C MeOH (86%)  Ts N O  2.45.1  2.45.2  H  same compound  Ts N O 2.45.2  2.14.4 Summary The acid catalyzed ring expansion reactions of allylic alcohols 2.30.2 and 2.30.6 were successful in both cases. Some of the products formed in the N-bromosuccinimide promoted ring expansion reactions were then used to confirm the configuration of products formed in the acid catalyzed semipinacol reactions. Substrates with groups at the 5- and 6-positions were found to give products that have the same relative configuration when either the Bronsted acid protocol or the N-bromosuccinimide protocol was used. In contrast, for substrates with groups at the 4-position, the relative configuration of the major product produced in the acid catalyzed ring expansion reaction was found to have the same relative configuration as the minor product formed in the N-bromosuccinimide promoted ring expansion reaction. 2.15 The Halichlorine Synthesis Revisited 2.15.1 Introduction The N-bromosuccinimide protocol for making 6-azapirocyclic ketones was explored with the intention of using one of the products in the total synthesis of halichlorine. Specifically compounds 2.36.8a and 2.44.1 were targeted (Figure 2. 13). The ketone functionality of these compounds was meant to be used as a means of elaborating the C14 to C21 side chain of halichlorine. The group at C13 of the C ring and the group attached to C5 of the B ring of halichlorine have a cis relationship. The corresponding groups in compounds 2.36.8a and 2.29.1 are trans and therefore these compounds would not be suitable for the halichlorine synthesis  110  Chapter 2 A First Generation Approach Towards Halichlorine…  project. A possible method to obtain the desired cis relationship would be to exchange the ethoxy group of spirocyclopentanone 2.29.1 for an allyl group. Figure 2. 13 Potential Intermediates for the Halichlorine Synthesis H  trans  Ts O  A  O  5  N  21  Br  O  13  2.36.8a  Cl  D  5  N  B  H  cis  C  14 13  HO halichlorine  or  EtO ?  5  trans  Ts O  N  Br  cis  13  Ts O  N  Br  2.29.1  2.15.2 Synthetic Elaboration of 2.29.1 Unfortunately, efforts to substitute the ethyl aminal function with an allyl group using allyltrimethylsilane and a Lewis acid resulted in the formation of alkene 2.46.1 (Scheme 2. 46). Interestingly, an allyl group could be installed at the 6-position of the heterocycle using alkene 2.46.1. Reacting 2.46.1 with allyltrimethylsilane and trifluoroacetic acid generated bromide 2.36.8a as a single diastereomer. 76 Note that this compound has the same structure as the compound generated from the ring expansion of 6-allyl substrate 2.16.2 which unfortunately has the undesirable trans relationship between the allyl group and the ketone.  Scheme 2. 46 Attempted Elaboration of Compound 2.29.1 EtO Ts  H  H SiMe 3  SiMe 3 N  2.29.1  Br O  MgBr2. OEt2 (90 %)  Ts  N  2.46.1  Br O  CF3CO2H (51 %)  Ts  N  2.36.8a  Br O  111  Chapter 2 A First Generation Approach Towards Halichlorine…  2.15.3 Siloxy-epoxide Approach Towards Halichlorine Previous work in the Dake group involved semipinacol rearrangement reactions of siloxy-epoxides. 77 This protocol was incorporated in a formal total synthesis of fasicularin; this work was mentioned briefly earlier in this chapter (See Scheme 2. 5). This methodology starts with allylic cycloalkanols similar to those used in the acid catalyzed and N-bromosuccinimide promoted ring expansion reactions. A three step sequence was incorporated which involves: a) protection of the allylic alcohol as its silyl ether b) epoxidation of the alkene and c) Lewis acid promoted ring expansion to provide hydroxy azaspirocyclicketones. An example is given in Scheme 2. 47 where allylic cyclopentanol 2.30.4 is converted into hydroxy azaspirocyclohexanone 2.5.12. In this case the product was isolated as a single diastereomer; the major product was the one resulting from antiperiplanar migration of the alkyl group. Interestingly, the corresponding siloxy-epoxide semipinacol rearrangement of 2.47.2a was much less selective resulting in a ~ 1:1 mixture of spirocyclic ketones 2.47.3 and 2.47.4 (Scheme 2. 47). Note that in compound 2.47.3 there is a trans relationship between the hydroxyl group and the ketone. Presumably this product is the result of syn migration of the alkyl group. Scheme 2. 47 Siloxy-epoxide Semipinacol Rearrangements 5-Membered Ring OTBS  Ts  TMSOTf 2,6-lutidine THF, rt  N  Ts  DMDO, K2CO3 (98 %)  N  OH  TiCl4, CH2Cl2 Ts  N O OTMS  OTMS  2.30.4  OTBS  OTBS  OTBS  2.47.1  2.5.1  -78 oC, (96 %)  Ts  N  OH O  2.5.2  4-Membered Ring OTBS  Ts  OTBS TiCl4, CH2Cl2  N O OTMS 2.47.2a  -78 oC, (95 %)  Ts  N  OH O  2.47.3(50 %)  OTBS  +  Ts  N  OH O  2.47.4 (45 %)  It was later discovered that the diastereoselectivity of the siloxy-epoxide semipinacol rearrangement of 4-membered rings to 5-membered rings could be altered to favour the  112  Chapter 2 A First Generation Approach Towards Halichlorine…  formation of one diastereomer over the other (Table 2. 14). The diastereoselectivity was found to be dependant upon the choice of silyl protecting group on the alcohol as well as the choice of Lewis acid used to promote the rearrangement reaction. When trimethylsilyl (TMS) was used as the silyl protecting group and ytterbium (III) triflate was used as the Lewis acid promoter, the diastereomer in which the ketone and the hydroxyl group are cis was formed as the major product (entry 1). Alternatively, when triisopropylsilyl (TIPS) was used as the protecting group and titanium (IV) chloride was used as the Lewis acid promoter the diastereomer in which the ketone and the hydroxyl group are trans was formed as the major product (entry 2). Because both diastereomers are accessible in a reasonably selective manner, these processes are complimentary. Table 2. 14 Complimentary Diastereoselectivity in the Siloxy-epoxide Semipinacol Rearrangement OTBS  Ts  N O OR  OTBS  Lewis acid CH2Cl2  2.47.2  Ts  N  OH O  OTBS  +  Ts  2.47.3  Entry Substrate  R  OH  N  O  2.47.4  Lewis acid Temp. (°C) Yield (%) Ratio (2.47.3:2.47.4)  1  2.47.2a  TMS  Yb(OTf)3  0  87  4.4:1  2  2.47.2b  TIPS  TiCl4  -78  94  1:6.2  The siloxy-epoxide strategy described above might be applicable in the halichlorine synthesis. The N-bromosuccinimide promoted rearrangement reaction of allylic alcohol 2.16.2 provided spirocyclopentanone 2.36.8a as a single diastereomer. As previously mentioned, spirocyclopentanone 2.36.8a has a cis relationship between the bromine and the ketone. It was hoped that if the siloxy-epoxide strategy were applied to allylic alcohol 2.16.2 that the major product would be the one where there is a trans relationship between the hydroxyl group and the ketone. This would necessarily place the ketone cis to the allyl group. At the time these experiments were conducted we were in possession of significant amounts of allylic alcohol 2.30.6 i.e. the enantiomer of allylic alcohol 2.16.2 required for the halichlorine synthesis. Rather than waste the desired compound 2.16.2 on test reactions the siloxy-epoxide strategy was explored using allylic alcohol 2.30.6. The tertiary alcohol of 2.30.6 was protected as its triisopropylsilyl (TIPS) ether by treating with triisopropylsilyl  113  Chapter 2 A First Generation Approach Towards Halichlorine…  trifluoromethanesulfonate and 2,6-lutidine (Scheme 2. 48). Epoxidation of 2.48.1 gave an intermediate epoxide 2.48.2 which when treated with titanium (IV) chloride resulted in the formation of 6-azaspirocycle 2.48.3. The epoxide could not be isolated in pure form as it underwent spontaneous ring expansion upon concentration. Unfortunately spirocyclopentanone 2.48.3 has the undesired cis relationship between the hydroxyl group and the ketone.  Scheme 2. 48 Siloxy-Epoxide Rearrangement of 2.30.6  Ts  H  H  H N OH  TIPSOTf, 2,6-lutidine CH2Cl2 (38 %)  2.30.6  Ts  m-CPBA, NaHCO3  N OTIPS  CH2Cl2, H2O (80 %)  2.48.1 TiCl4  CH2Cl2, 4 A MS (60 %)  Ts  N O OTIPS  2.48.2  H Ts N  OH O  2.48.3  The configuration of alcohol 2.48.3 was verified though chemical correlation with bromide 2.36.7b, the product formed in the N-bromosuccinimide promoted ring expansion of allylic alcohol 2.30.6 (Scheme 2. 49). The secondary alcohol of 2.48.3 was activated as its methanesulfonate derivative 2.49.1. This compound was heated in toluene in the presence of 1,8-diazabicyclo-[5.4.0]-undec-7-ene (DBU) that resulted in an E2 elimination reaction to provide diene 2.45.1. A similar E2 elimination was carried out with bromide 2.36.7b (see Scheme 2.45 above). This resulted in the formation of a compound whose spectral data was identical with the compound formed here. Obviously, the siloxy-epoxide approach did not provide the desired outcome. The ketone and the allyl group are trans which means that this method cannot be used in the total synthesis of halichlorine.  114  Chapter 2 A First Generation Approach Towards Halichlorine…  Scheme 2. 49 Chemical Correlation of Compound 2.48.3 with Compound 2.36.7b H  H Ts N  OH O  MsCl, DMAP CH2Cl2 (89 %)  2.48.3  OMs O  DBU, PhCH3 reflux (23 %)  2.49.1  H  Ts  Ts N  H  i  N  NBS O PrOH, o  OH  -78 C (98 %)  2.30.6  Ts N O 2.45.1  H Ts N  Br O  DBU, PhCH3 reflux (27 %)  2.36.7b  2.16 Chapter Summary In this chapter, two possible pathways towards the total synthesis of halichlorine were examined. In Path A, a new ring closing metathesis catalyst was introduced that was subsequently used to successfully close the A ring of halichlorine. However, subsequent transformations in this pathway did not prove successful. In Path B it was discovered that acid could not be used to promote the ring expansion of allylic alcohol 2.1.10. This unsuccessful result led to the discovery that N-bromosuccinimide could be used to promote the semipinacol rearrangement of a series of piperidine-derived allylic cyclobutanols to the corresponding 6-azaspirocyclopentanones. The yields for these reactions were uniformly high and in some cases were completely diastereoselective. In all cases the products that were formed were the result of migration of the alkyl group anti to the intermediate bromonium ion. Consequently the products were always those in which the bromine and the ketone have a cis relationship. When N-bromosuccinimide was used to promote the ring expansion of a 5-membered ring to a 6-membered ring the reaction failed. The acid catalyzed ring expansion reactions of the 6-allyl and the 4-isopropyl substrates were also done. Afterwards the products formed in the acid catalyzed ring expansion reactions were chemically correlated to those compounds formed in the N-bromosuccinimide promoted ring expansion reactions. Substrates with groups at the 5- and 6-positions were found to react similarly under both sets of conditions while substrates with groups at the 4-position were found to react with the opposite relative configuration under each of the reaction conditions. Ultimately none of the products obtained in the N-bromosuccinimide  Chapter 2 A First Generation Approach Towards Halichlorine…  115  rearrangement reaction were useable in the total synthesis of halichlorine. This was because the key functional groups exhibited a trans relationship whereas halichlorine required these functional groups to be cis. Finally, an attempt to use the siloxy-epoxide rearrangement to favour the desired “cis” diastereomer was not successful. Clearly, a different approach would be needed if the synthesis of halichlorine were to be achieved.  Chapter 2 A First Generation Approach Towards Halichlorine…  116  2.17 Experimental Unless otherwise stated, all reactions were performed under a nitrogen atmosphere in flame-dried glassware. The glass syringes, Teflon® cannulae and stainless steel needles used for handling anhydrous solvents and reagents were oven dried, cooled in a dessicator, and flushed with dry nitrogen prior to use. Plastic syringes were flushed with dry nitrogen before use. Thin layer chromatography (TLC) was performed on DC-Fertigplatten SIL G-25 UV254 pre-coated TLC plates. Gas chromatographic (GC) analyses in a helium carrier gas were performed on an Agilent 6890 gas chromatograph, equipped with an Agilent 5973N mass selective detector. A Hewlett-Packard HP-5MS (30 m × 0.25 mm × 0.25 μm ID) fused silica capillary columns was used. Melting points were performed using a Mel-Temp II apparatus (Lab devices USA) and are uncorrected. Optical rotations of samples were performed using a Jasco model P1010 polarimeter at 589 nm (sodium ‘D’ Line). Infrared (IR) spectra were obtained using a PerkinElmer FT-IR spectrometer. Proton nuclear magnetic resonance (1H NMR) and carbon nuclear magnetic resonance (13C NMR) spectra were recorded in deuterochloroform (unless otherwise indicated) using either a Bruker AV-300, a Bruker WH-400 or a Bruker AV-400 spectrometer. Chemical shifts are reported in parts per million (ppm) and are referenced to the centreline of deuterochloroform (δ 7.24 ppm 1H NMR, 77.0 ppm 13C NMR). Coupling constants (J values) are given in Hertz (Hz). Low resolution mass spectra (LRMS) were recorded on either an Agilent 5973N mass selective detector, attached to an Agilent 6890 gas chromatograph for electron impact ionization (EI), a Kratos MS 80 spectrometer for desorption chemical ionization (CI) with the ionization gas noted, or a Agilent HP1100 spectrometer for electrospray ionization (ESI). Microanalyses were performed by the Microanalytical Laboratory at the University of British Columbia on a Carlo Erba Elemental Analyzer Model 1106 or a Fisions CHN-O Elemental Analyzer Model 1108. All solvents and reagents were purified and dried using established procedures. 78 Tetrahydrofuran, 1,2-dimethoxyethane and diethyl ether were distilled from sodium benzophenone ketyl under an atmosphere of dry argon. Methylene chloride, triethylamine, dimethyl sulfoxide, acetonitrile, pyridine, 2,6-lutidine, N,N-dimethylacetamide and dimethylsulfoxide were distilled from calcium hydride over an atmosphere of dry argon or dry nitrogen. Trifluoromethanesulfonic anhydride was distilled from phosphorous pentaoxide over an atmosphere of dry nitrogen. N,N-Dimethylformamide was dried by storing over 4 Å molecular sieves three times over three successive days. Solutions of n-butyllithium or methyllithium in hexanes were obtained from the Aldrich Chemical Co. and were standardized  Chapter 2 A First Generation Approach Towards Halichlorine…  by titration against diphenylacetic acid.  117  Elemental zinc was purified and made active prior to  use by treatment with hydrochloric acid according to the method of Hauser and Breslow. 79 Unless otherwise indicated all other reagents were commercially available and were used without further purification. (±)-6-Ethoxy-piperidine-2-one (2.6.1)2a  O H  NaBH4 HCl, EtOH  N  EtO  O 2.1.7  H  H N O 2.6.1  This reaction was performed open to the atmosphere. To a rapidly stirred solution of 4.74 g of glutarimide (2.1.7) (42.9 mmol) in 250 mL of ethanol at 0 °C was added 3.63 g of sodium borohydride (95.9 mmol). As the reaction mixture was stirred at 0 °C, four (4) drops of a solution of 2M hydrochloric acid in ethanol was added to the reaction mixture every fifteen min for 5 h. A solution of 2M hydrochloric acid in ethanol was added until the pH of the solution was 3. The reaction mixture was stirred for 2 h. Neutralization of the reaction through the addition of 3% potassium hydroxide in ethanol produced a milky white solution. After removal of solvent in vacuo, the resulting solid was washed with dichloromethane (3x100 mL). The combined organic extracts were filtered, dried over magnesium sulfate, filtered and concentrated in vacuo. Purification by column chromatography on silica gel (80% ethyl acetate-hexanes) gave 3.92 g (72%) of a white powder. The spectral data given below match those described in reference 2a. mp=122-123 °C. IR: 3190, 2972, 1670 cm-1; 1H NMR (400 MHz, CDCl3): δ 6.62 (s, 1H), 4.66 (dd, J=7.0 Hz, 3.4 1H), 3.61 (dq, J=8.9, 7.0 Hz, 1H), 3.42 (dq, J=8.9, 7.0 Hz, 1H), 2.47-2.25 (m, 2H), 2.11-1.85 (m, 2H), 1.83–1.64 (m, 2H), 1.21 (t, J=7.0 Hz, 3H). (±)-6-allylpiperidin-2-one (2.6.2)  EtO H  H  H TMS  N  BF3 .OEt2 O  2.6.1  H  N O  2.6.2  Chapter 2 A First Generation Approach Towards Halichlorine…  118  To a stirred solution of 5.00 g of 6-ethoxy-2-piperidin-2-one (2.6.1) (34.9 mmol, 1 equiv) in 200 mL of dichloromethane at 0 °C was added 13.9 mL of allyltrimethylsilane (10.0 g, 87.3 mmol, 2.5 equiv) was added and the resulting solution was stirred at 0 °C for 5 min. Boron trifluoride·diethyl etherate (8.85 mL, 9.91 g, 69.8 mmol, 2 equiv) was added and the resulting solution was stirred at rt for 20 h. Brine (200 mL) was added and the layers were separated. The organic layer was dried over magnesium sulfate, filtered and solvents were removed in vacuo. Purification by column chromatography on silica gel (5 % methanol-methylene chloride) gave 4.86 g (quantitative) of a white solid. The spectral data given below match those described in reference 24. IR (KBr): 3192, 1774 cm-1. 1H NMR (400 MHz, CDCl3): δ 5.82 (bs, 1H), 5.76-5.64 (m, 1H), 5.18-5.10 (m, 2H), 3.41-3.33 (m, 1H), 2.42-2.20 (m, 3H), 2.14-2.04 (m, 1H), 1.93-1.83 (m, 2H), 1.73-1.61 (m, 1H), 1.41-1.30 (m, 1H). (±)-1,6-diallylpiperidin-2-one (2.1.6a)  H  H NaH,  H  N  Br N  O 2.6.2  O 2.1.6a  Small Scale A solution of 86.4 mg of sodium hydride (2.16 mmol, 1.2 equiv, 60 % suspension in mineral oil (Aldrich)) in 5 mL of THF was cooled to 0 °C. A solution of 250 mg of 6-allyl-2-piperidin-2one (2.6.2) (1.80 mmol, 1 equiv) in 5 mL of THF was added and the resulting mixture was stirred 0 °C for 1h. Allyl bromide (187 μL, 261 mg, 2.16 mmol, 1.2 equiv) was added and the reaction mixture was stirred at rt for 3 d. As the reaction had failed to go to completion, sodium hydride (86 mg, 2.16 mmol, 1 equiv) was added, the reaction was stirred at rt for 10 min, allyl bromide (187 μL, 2.16 mmol, 1.2 equiv) was added and the resulting mixture was stirred at rt for 1 d. After removing the solvent in vacuo, water (5 mL) and 5 mL of dichloromethane were added and the layers were separated. The aqueous layer was extracted with dichloromethane (2x5 mL). The combined organic layers were dried over magnesium sulfate, filtered and solvents were removed in vacuo. Purification by column chromatography on silica gel (15 % diethyl ether-dichloromethane) gave 312 mg (97 %) of a yellow oil.  Chapter 2 A First Generation Approach Towards Halichlorine…  119  Larger Scale The reaction was carried out as above using 690 mg of sodium hydride (17.2 mmol, 1.2 equiv, 60 % dispersion in mineral oil (Aldrich)), 2.00 g of 6-allyl-2-piperidin-2-o