Copyright by Benjamin Perry Fauber 2006 · The first total synthesis of galanthamine was reported...

331
Copyright by Benjamin Perry Fauber 2006

Transcript of Copyright by Benjamin Perry Fauber 2006 · The first total synthesis of galanthamine was reported...

  • Copyright

    by

    Benjamin Perry Fauber

    2006

  • The Dissertation Committee for Benjamin Perry Fauber Certifies that this is the

    approved version of the following dissertation:

    STUDIES DIRECTED TOWARD THE SYNTHESES OF THE

    BIOLOGICALLY ACTIVE ALKALOIDS

    (-)-GALANTHAMINE AND (-)-LEMONOMYCIN

    Committee:

    Philip D. Magnus, Supervisor

    Richard A. Jones

    Sean M. Kerwin

    Michael J. Krische

    Hung-wen Liu

  • STUDIES DIRECTED TOWARD THE SYNTHESES OF THE

    BIOLOGICALLY ACTIVE ALKALOIDS

    (-)-GALANTHAMINE AND (-)-LEMONOMYCIN

    by

    Benjamin Perry Fauber, B.S.

    Dissertation

    Presented to the Faculty of the Graduate School of

    The University of Texas at Austin

    in Partial Fulfillment

    of the Requirements

    for the Degree of

    Doctor of Philosophy

    The University of Texas at Austin

    December, 2006

  • Dedication

    Laura, Richard, Jill, Dan, Frank, and Sallie — for all of your support and encouragement

  • v

    Acknowledgements

    I am indebted to Professor Philip D. Magnus for the mentoring and guidance he

    provided in solving the complex problems associated with my projects. Additionally, I

    would like to thank Dr. Vincent Lynch for his assistance with X-ray crystallography and

    structural elucidation, as well as the members of the P. D. Magnus and S. F. Martin

    research groups for their helpful insight and discussions.

  • vi

    STUDIES DIRECTED TOWARD THE SYNTHESES OF THE

    BIOLOGICALLY ACTIVE ALKALOIDS

    (-)-GALANTHAMINE AND (-)-LEMONOMYCIN

    Publication No._____________

    Benjamin Perry Fauber, Ph.D.

    The University of Texas at Austin, 2006

    Supervisor: Philip Douglas Magnus

    Despite the enormous amount of work devoted to the synthesis of the

    anticholinesterase Amaryllidaceae alkaloid, (-)-galanthamine, and the diversity of the

    various strategies employed, the para-alkylation of an appropriately substituted phenol to

    generate the cross-conjugated 2,4-cyclohexadienone has not been reported. As discussed

    in this dissertation, the successful implementation of the intramolecular phenolate

    alkylation strategy avoids the low yielding phenolic oxidation reaction used previously to

    generate similar intermediates. The resultant product requires a reductive amination of

    the aromatic aldehyde and latent aliphatic aldehyde to arrive at racemic narwedine, a

    biogenetically related and validated synthetic precursor to (-)-galanthamine.

    A methodology for the construction of enantio-enriched hydroisoquinolines was

    also developed, with potential application toward the synthesis of the

    tetrahydroisoquinoline antitumor antibiotic (-)-lemonomycin. Several approaches are

  • vii

    discussed, with the key step being an asymmetric reduction of 1-substituted and

    1,3-disubstituted isoquinolines to yield enantio-enriched hydroisoquinoline products.

  • viii

    Table of Contents

    List of Abbreviations ...............................................................................................x

    PART A: STUDIES DIRECTED TOWARD THE SYNTHESIS OF THE BIOLOGICALLY ACTIVE ALKALOID (-)-GALANTHAMINE 1

    Chapter 1: The Amaryllidaceae Alkaloids and Galanthamine ................................2

    1.0 Introduction...............................................................................................2

    1.1 Galanthamine ............................................................................................6

    1.2 References...............................................................................................16

    Chapter 2: Studies Toward the Synthesis of (-)-Galanthamine .............................21

    2.0 Double Condensation Strategy ...............................................................21

    2.1 Benzoazepine Strategy............................................................................24

    2.2 Pummerer Strategy..................................................................................33

    2.3 Ether Alkylation Strategy .......................................................................38

    2.4 Amination of the Cross-Conjugated Dienone.........................................51

    2.5 References...............................................................................................64

    PART B: STUDIES DIRECTED TOWARD THE SYNTHESIS OF THE BIOLOGICALLY ACTIVE ALKALOID (-)-LEMONOMYCIN 73

    Chapter 3: The Tetrahydroisoquinoline Alkaloids and Lemonomycin .................74

    3.0 Introduction.............................................................................................74

    3.1 Lemonomycin and Prior Synthetic Studies.............................................79

    3.2 Previous Work within the Magnus Research Group...............................84

    3.3 Summary of Previous Efforts Toward Lemonomycin............................87

    3.4 References...............................................................................................89

    Chapter 4: Studies Toward the Synthesis of (-)-Lemonomycin ............................94

    4.0 Background on the Asymmetric Synthesis of Isoquinolines ..................94

    4.1 Asymmetric Reduction Strategy .............................................................97

    4.2 Bischler-Napieralski Strategy ...............................................................118

    4.3 References.............................................................................................128

  • ix

    PART C: EXPERIMENTAL CONDITIONS AND COMPOUND DATA 138

    Chapter 5: Experimental Conditions and Compound Data..................................139

    5.0 General Information..............................................................................139

    5.1 Experimental Conditions and Compound Data for Chapter 2 ..............141

    5.2 Experimental Conditions and Compound Data for Chapter 4 ..............184

    5.3 References.............................................................................................232

    Appendix A: X-ray Data for the Biaryl Dimer (36) ............................................234

    Appendix B: X-ray Data for the Cross-Conjugated Dienone (79) ......................241

    Appendix C: X-ray Data for the Tetracyclic Pyranone (84)................................247

    Appendix D: X-ray Data for the Multicyclic Tertiary Amine (93) .....................255

    Appendix E: X-ray Data for the Lactone Enone (103) ........................................261

    Appendix F: X-ray Data for the Aldehyde Lactol (105)......................................269

    Appendix G: X-ray Data for the Acetal Ether (107)............................................277

    Appendix H: X-ray Data for the Isoquinoline Amide (39)..................................283

    Appendices References........................................................................................288

    Bibliography ........................................................................................................290

    Vita.......................................................................................................................317

  • x

    List of Abbreviations

    1,2-DCE 1,2-dichloroethane

    1,2-DME 1,2-dimethoxyethane

    Å ångström

    Ac acetyl

    Ar aryl or argon (dependent upon the context)

    atm atmosphere

    BHT 2,6-di-tert-butyl-4-methyl-phenol

    binap 2,2'-bis(diphenylphosphino)-1,1'-binaphthyl

    biphep 2,2'-bis(diphenylphosphino)-1,1'-biphenyl

    Bn benzyl

    Boc tert-butylcarbonyl

    bs broad singlet

    Bu butyl

    c concentration

    CAN ammonium cerium(IV) nitrate

    Cbz benzyloxycarbonyl

    m-CPBA m-chloroperbenzoic acid

    cod 1,5-cyclooctadiene

    CSA 10-camphorsulfonic acid

    Cy cyclohexyl

    d doublet

    D sodium D-line, 589 nm

  • xi

    DABCO 1,4-diazabicyclo[2.2.2]octane

    dba dibenzylideneacetone

    DBU 1,8-diazabicyclo[5.4.0]undec-7-ene

    DCC N,N'-dicyclohexylcarbodiimide

    DCE dichloroethane

    DCM dichloromethane

    dd doublet of doublets

    DEAD diethyl azodicarboxylate

    DIAD diisobutyl azodicarboxylate

    DIBAL-H diisobutylaluminum hydride

    DIEA N,N-diisopropylethylamine

    diop 4,5-bis(diphenylphosphinomethyl)-2,2-dimethyl-1,3-dioxolane

    dioxane 1,4-dioxane

    DMAP 4-N,N-dimethylaminopyridine

    DMF N,N-dimethylformamide

    DMP Dess-Martin periodinane

    DMSO dimethyl sulfoxide

    dppp 1,3-bis(diphenylphosphino)propane

    dr diastereomeric ratio

    dt doublet of triplets

    EDC N-ethyl-N'-(3-dimethylaminopropyl)carbodiimide

    ee enantiomeric excess

    er enantiomeric ratio

    Et ethyl

    g gram

  • xii

    Gly glycosyl

    HOBt 1-hydroxybenzotriazole

    HPLC high-pressure liquid chromatography

    HRMS high-resolution mass spectrometry

    IR infrared

    kg kilogram

    L-selectride lithium tri-sec-butylborohydride

    LDA lithium diisopropylamide

    m multiplet

    m meta

    M molar

    Me methyl

    ml milliliter

    mmol millimole

    mol mole

    MOM methoxymethyl

    M.p. melting point

    Ms methanesulfonyl

    MTBE methyl-tert-butyl ether

    NBS N-bromosuccinimide

    ng nanogram

    nm nanometer

    NMR nuclear magnetic resonance

    o ortho

    p para

  • xiii

    Ph phenyl

    Piv pivaloyl

    ppm parts per million

    PPSE polyphosphoric acid trimethylsilyl ester

    Pr propyl

    prep preparatory

    psi pounds per square inch

    py pyridine

    q quartet

    qd quadruplet of doublets

    R generic functional group

    s singlet

    t triplet

    TBAF tetrabutylammonium fluoride

    TBDMS tert-butyldimethylsilyl

    td triplet of doublets

    Tf trifluoromethanesulfonyl

    TFA trifluoroacetic acid

    TFAA trifluoroacetic anhydride

    THF tetrahydrofuran

    THP tetrahydropyran

    TIPS triisopropylsilyl

    TLC thin-layer chromatography

    TMS trimethylsilyl

    Troc 2,2,2-trichloroethoxycarbonyl

  • xiv

    Ts toluenesulfonyl

    p-TsOH p-toluenesulfonic acid

    X halogen

  • 1

    PART A: STUDIES DIRECTED TOWARD THE

    SYNTHESIS OF THE BIOLOGICALLY ACTIVE

    ALKALOID (-)-GALANTHAMINE

  • Chapter 1: The Amaryllidaceae Alkaloids and Galanthamine

    1.0 INTRODUCTION

    The Amaryllidaceae alkaloids comprise a unique group of tertiary amine bases

    which have been found almost exclusively within the title family. The Amaryllidaceae

    family includes (but is not limited to) the Amaryllis, Crinum, Galanthus, Haemanthus,

    Hymenocallis, Hypoxis, Leucojum, Lophiola, Lycorus, Narcissus, Sceletium, Vallota, and

    Zephyranthes genera.1,2 Members of this family are found terrestrially, throughout the

    globe, with the plants being most prevalent in tropical and subtropical regions.3 The

    alkaloids isolated from plants of the Amaryllidaceae family are often most abundant in

    the buds and bulbs, giving rise to 0.1-2% total alkaloid content versus raw plant

    material.4

    N

    O

    O

    HO

    lycorine (1)

    O

    OH

    O

    N

    galanthamine (3)

    O

    O N

    OH

    crinine (2)

    membrenone (4)

    NH

    O

    O

    pancratistatin (6)

    O

    OHOH

    OHO

    O N

    O

    OH

    montanine (5) OHN

    HO

    OO HO

    Figure 1.01. Some representative Amaryllidaceae alkaloids

    2

  • 3

    The Amaryllidaceae alkaloid family is distinguished by the elaboration of a

    15 carbon nucleus ― of which there is typically an aromatic unit and a reduced aromatic

    unit.5 The greater than 500 alkaloids found in the family can be attributed to the diversity

    that results from minor variations in the degree of oxidation, aromatic substitution,

    hydrogenation, and ring connectivity of the parent C-15 skeleton (Figure 1.01). 6,7

    The similarities of the Amaryllidaceae alkaloids are best rationalized through their

    biosynthetic pathways. The alkaloids arise via alternative modes of norbelladine (9)

    oxidative coupling. Norbelladine is generated through the reductive condensation of

    3,4-dihydroxybenzaldeyde (7) with tyramine (8), and both of these precursors are the

    products of phenylalanine and tyrosine metabolites, respectively. Three structural types

    can be generated through the different modes of phenolic oxidation coupling of 4'-O-

    methylnorbelladine (10) ― para-ortho' (A), para-para' (B), and ortho-para' (C) (Scheme

    1.01). These three oxidative coupling modes give rise to the respective skeletal

    frameworks of lycorine (1), crinine (2), and galanthamine (3).8 Subsequent redox of the

    crinine core results in the formation of the membrenone (4), montanine (5), and

    pancratistatin (6) skeletons.

  • N

    O

    O

    HO

    lycorine (1)

    O

    OH

    O

    Ngalanthamine (3)

    O

    O N

    OH

    crinine (2)

    L-PheHO

    HOO + NH2 OH

    L-Tyr

    tyramine (8)

    HO

    HONH

    norbelladine (9)

    S-methyl adenosine

    O

    HOHN

    HO

    OOH

    NH

    OH

    O

    NH

    OH

    HO

    (7)

    (10) (10)(10)

    B CA

    OH

    reductive condensation

    Scheme 1.01. Biosyntheses of the Amaryllidaceae alkaloid skeletons

    The medicinal properties of the Amaryllidaceae family have long been exploited

    by natives of the high-mountain Caucasus region of Russia, where teas made from the

    plant’s constituents are used to treat a variety of ailments.9 Modern research efforts have

    revealed the individual Amaryllidaceae alkaloid components to be potent medicinal

    agents, with a wide range of biological activity. Noteworthy examples include the highly

    explored antineoplastic pancratistatin (6), which inhibits murine P388 lymphomic

    leukemia and human cancer cell lines with double-digit ng/ml activity.10

    4

  • 5

    Membrenone (4), and its oxidative related derivatives, have received notable synthetic

    activity due to their action as serotonin uptake inhibitors.7 Additionally,

    (-)-galanthamine (3) has been thoroughly explored as a symptomatic treatment for

    Alzheimer’s disease, due to its activity as an acetylcholinesterase inhibitor. The

    following subchapter discusses the background, synthetic studies, and biological activity

    of galanthamine in more detail.

  • 1.1 GALANTHAMINE

    The alkaloid (-)-galanthamine (3) is a tertiary amine base which has been isolated

    from a number of species in the Amaryllidaceae family. Galanthamine was first reported

    as a constituent of the Caucasian snowdrop (Galanthus woronowii) in 1952.11 Its

    structure was confirmed, utilizing degradation studies, by Uyeo and Kobayashi in 1957.9

    The absolute configuration of the structure was determined after Barton and Kirby

    reported the first total synthesis of (-)-galanthamine in 1960.12

    Several genera produce galanthamine, with most yielding approximately

    10-100 mg of galanthamine per kilogram of raw plant material through a costly

    extraction process (approximately US$50,000 per kilogram of galanthamine).13 Due to

    the high cost of extraction, several synthetic routes to galanthamine, and the related

    oxidized enone, narwedine (11), have been explored (Scheme 1.02).

    O

    ON

    O

    O

    O

    OH

    N

    (-)-galanthamine (3) (-)-narwedine (11)

    O

    O

    OH

    N

    (-)-lycoramine (12)

    Scheme 1.02. The structures of (-)-galanthamine, (-)-narwedine, and

    (-)-lycoramine

    The first total synthesis of galanthamine was reported by Barton and Kirby in

    1960.12 They completed a biomimetic route, in which 4'-O-N-dimethylnorbelladine (13)

    was subjected to phenolic oxidation coupling conditions, giving rise to a symmetrical

    dienone intermediate (14), which was then trapped by the adjacent phenol, yielding

    6

  • (±)-narwedine (11), the proposed biogenetic precursor to galanthamine. The carbonyl of

    narwedine was reduced with LiAlH4 to yield a mixture of (±)-galanthamine (3), and

    (±)-epigalanthamine (15) (Scheme 1.03).

    O

    ON

    O

    O

    O

    OH

    N

    (±)-(11)

    OOH

    N

    OH

    (13)

    O

    O

    OH

    N+

    HO

    ON

    O

    (14)

    K3[Fe(CN)6]

    NaHCO3H2O/CHCl30.9% yield

    LiAlH4

    Et2Oreflux

    (±)-(3)39% yield

    (±)-(15)39% yield

    Scheme 1.03. Barton and Kirby’s synthesis of (±)-galanthamine

    While the yield of the phenolic oxidation step was 0.9%, the Barton and Kirby

    approach demonstrated the plausibility of utilizing a phenolic oxidation strategy to

    generate the galanthamine core. Additionally, feeding experiments involving 14C-labeled

    4'-O-N-dimethylnorbelladine, and subsequent incorporation of the labeled carbon unit

    into galanthamine and narwedine, further supported the hypothesis for the biosynthesis of

    galanthamine from norbelladine (9) intermediates, just as Barton and Kirby had used in

    their synthesis.14

    Zenk and co-workers reported a detailed study on the biosynthesis of

    galanthamine and found that 13C-labeled 4'-O-methylnorbelladine was 27% incorporated

    into galanthamine, and 31% incorporated into N-demethylgalanthamine (18), through

    7

  • feeding experiments with the Snowflake Lily (Leucojum aestivum).15 Furthermore, they

    established that N-demethylgalanthamine was N-methylated in the final step of the

    biosynthesis, which was in contrast to the biosynthetic hypothesis of Barton and Kirby.

    Thus, Zenk and co-workers concluded that norbelladine (9) reacted within the Snowflake

    Lily to form N-demethylnarwedine (17), which was reduced to N-demethylgalanthamine

    (18), and N-methylated to form galanthamine (3) (Scheme 1.04).

    O

    ONH

    O

    O

    O

    OH

    NH

    (17)

    OOH

    NH

    OH

    (9)

    HO

    ONH

    O

    (16)

    (18)

    O

    O

    OH

    N

    (-)-(3)

    Scheme 1.04. Zenk’s postulated biosynthetic pathway for galanthamine

    Interestingly, in a process first reported by Barton and Kirby in their 1962 full

    paper detailing the synthesis of (-)-galanthamine (3), (–)-narwedine (11) was converted,

    in high optical purity, to (+)-narwedine by trace amounts of (-)-galanthamine during

    recrystallization from hot ethanol (Scheme 1.05).16 This method was exploited to

    generate (-)-narwedine, by resolving (±)-narwedine with catalytic amounts of

    (+)-galanthamine, which upon further reduction yielded (-)-galanthamine.

    8

  • O

    ON

    O

    (-)-(11)

    O

    O

    OH

    N

    (-)-(3)

    MnO2

    CHCl323 °C

    EtOH recrx.O

    ON

    O

    (+)-(11)

    trace of(-)-galanthamine

    80% yield38% yield

    Scheme 1.05. Barton and Kirby’s resolution of (–)-narwedine to

    (+)-narwedine

    The dynamic resolution of narwedine was improved upon by the scientists at

    Ciba-Geigy, utilizing catalytic amounts of (+)-galanthamine to resolve (±)-narwedine

    (11) into (-)-narwedine in high optical yield and purity on kilogram scale (Scheme

    1.06).17

    O

    ON

    O

    O

    ON

    O

    (+)-galanthamine(1 mol %)

    95% EtOH/NEt3 (9:1)80 → 25 °C90% yield

    (±)-(11) (-)-(11)

    HO

    ON

    O

    (14)

    Scheme 1.06. The Ciba-Geigy dynamic resolution of (±)-narwedine

    The Ciba-Geigy team went on to explore the details of the resolution and found

    that no isolable, or spectrally identifiable, conglomerate or covalent pair of narwedine

    and galanthamine was detected in the solution or solid states. The Ciba group explained

    the resolution event as a “spontaneous resolution of stereoisomers,” in which the

    crystallization of a single enantiomer of (-)-narwedine drove the dynamic resolution

    event to completion through the symmetrical dienone intermediate (14). The scientists at 9

  • Ciba-Geigy also solved the problem of epigalanthamine (15) contamination during the

    reduction of narwedine to galanthamine by utilizing L-selectride as the reducing agent to

    generate (-)-galanthamine in 99% yield. It should also be noted that (±)-narwedine can

    be resolved via a classical chiral acid resolution with di-p-toluoyl-D-tartaric acid.18

    The syntheses that followed Barton and Kirby’s initial studies of galanthamine (3)

    have focused on generating the quaternary carbon center, while simultaneously forming

    the bond between the two six-membered rings of the target. The synthetic efforts to form

    the aforementioned carbon-carbon bond can be categorized into two different methods ―

    those which utilize a phenolic oxidation strategy and those which utilize a Heck

    coupling19 strategy (Scheme 1.07).13

    O

    O

    OH

    NO

    R'

    YPO

    XRphenolic oxidation Heck coupling

    (3) (20)

    O

    HON

    OH

    (19)X

    Scheme 1.07. The two major strategies to form the galanthamine core

    The biggest contribution to improving the yield of the phenolic oxidation

    coupling, over that of the 0.9% yield initially reported by Barton and Kirby, came from

    the research of Kametani and co-workers. They made use of a para-blocking group (in

    the form of a halide) (26), which promoted the desired ortho-para' coupling, to generate

    the galanthamine core (27) in 40% yield. Subsequent treatment with refluxing LiAlH4

    removed the aromatic halide blocking group and reduced the enone carbonyl to produce a

    mixture of (±)-galanthamine (3) and (±)-epigalanthamine (15) (Scheme 1.08).20

    10

  • O

    HON

    OH

    (26) BrO

    O

    ON

    O

    (±)-(27)

    K3[Fe(CN)6]

    NaHCO3H2O/CHCl3

    60 °C40% yield

    Br

    O

    ON

    OH

    BnO

    MeNH2NaOH

    H2O BnO

    HNCl

    O O91% yield

    LiAlH4

    THFreflux

    98% yieldBnO

    HN

    BnO

    O Br

    COCl

    10% NaOH(aq)CHCl3, 23 °C

    O

    BnON

    OBn

    (25) BrO

    (21) (22)(23)

    (24)

    48% HBr(aq)

    EtOHreflux

    55% yield(2 steps)

    LiAlH4

    THFreflux

    (±)-(3)50% yield

    O

    ON

    OH

    (±)-(15)40% yield

    +

    Scheme 1.08. Kametani’s synthesis of (±)-galanthamine

    The Kametani research group and others have also explored the effects of various

    phenolic oxidation reagents and substitution on the amine/amide tether, in attempts to

    improve the yield of the phenolic oxidation step (c.a. 0.9-60% yield).21 The effects of

    asymmetric centers on the amine tether, in efforts to promote formation of a single

    narwedine-type enantiomer, have also been examined.22

    Reports concerning the Heck coupling strategy to form the galanthamine core

    structure often involved the formation of a 2-haloarylether, which was then cyclized to

    yield the tricyclic galanthamine core. The research groups of Fels,23 Parsons,24 and

    Guillou/Thal25 independently reported this approach on nearly identical model substrates,

    which was built upon by the Trost research group.26 The Trost approach utilized an

    11

  • asymmetric allylic alkylation27 (30), followed by the Heck coupling step, to generate the

    tricyclic core of galanthamine as a single enantiomer (33) (Scheme 1.09).

    OO

    O

    CN

    O

    ON

    OH

    (-)-(3)

    (33)

    OO Br

    O

    (30)

    O

    OO Br

    O

    (32)

    CN

    OO Br

    O(31)

    OH

    O

    1) p-TsOH, (MeO)3CH MeOH

    2) DIBAL-H, -78 °C toluene 85% yield

    O

    1) Acetonecyanohydrin PPh3, DIAD, Et2O

    2) p-TsOH, H2O, THF76% yield96% ee

    Pd(OAc)2, dpppAg2CO3

    PhCH3107 °C

    91% yield

    SeO2NaH2PO4

    1,4-dioxane150 °C

    57% yield10:1 dr

    OO

    O

    CN

    (34)

    HO1) MeNH2, MeOH2) DIBAL-H3) NaH2PO4(aq)

    4) NaBH3CN

    O

    O

    BrOH O O

    OTroc+

    (28) (29)

    [(η3-C3H5)PdCl2]ligand*

    NEt3, DCM23 °C

    72% yield88% ee

    Scheme 1.09. Trost’s synthesis of (-)-galanthamine

    Although the yield for the Heck coupling reaction, which generated the

    quaternary carbon center, was higher than the respective phenolic oxidation reaction, the

    resulting spirocyclic cyclohexene product (33) required an allylic oxidation to install the

    allylic alcohol (34) present in galanthamine. Unfortunately, allylic oxidation reactions

    are often unselective,24 and the yields are less than 50%, as exemplified in the formation

    of the allylic alcohol product (34), highlighting a significant drawback to the Heck

    coupling approach. Moreover, the low yielding oxidation reaction brought the overall

    12

  • yield of the synthesis to a comparable level versus the phenolic oxidation strategy (c.a.

    10% overall yield).

    A recent publication from the Tu research group discussed a new approach to

    galanthamine,28 which was analogous to their earlier work toward lycoramine (12).29 The

    key step introduced the quaternary center of the target via a semi-pinacol rearrangement

    of a secondary alcohol (35). Electrophilic bromination of the olefin promoted the aryl

    migration, which quenched the bromonium ion, resulting in the formation of the bicyclic

    aldehyde (36). Removal of the silylether revealed the phenol, which displaced the

    secondary bromide under basic conditions, to yield the tricyclic core of

    galanthamine (37) (Scheme 1.10).

    OH

    O

    OO

    (35)

    NBS

    DCM0 °C

    95% yieldOTBDMS

    OO

    OTBDMSO

    O

    BrDBU

    DMSO95 °C

    90% yield

    (±)-(36)

    O

    O

    O

    (±)-(37)

    OO

    O

    O(±)-(38)

    O

    OO O

    O(±)-(39)

    O

    OO

    1) LDA, TMSCl THF, -78 °C

    2) Pd(OAc)2 Na2CO3 CH3CN

    63% yield

    O

    O(±)-(40)

    NHO

    HO

    O

    ON

    OH

    (±)-(3)

    LiAlH4

    1,2-DMEreflux

    76% yield

    (CH2O)nTFA

    1,2-DCE82% yield

    O

    ON

    OH

    (±)-(41)

    O

    Scheme 1.10. Tu’s synthesis of (±)-galanthamine

    13

  • 14

    Additional transformations homologated the aldehyde and revealed the ketone,

    generating the cyclohexanone intermediate (38). The oxidation of the ketone to the

    enone (39) proceeded in low yield, highlighting a drawback to this approach. Subsequent

    reduction of the enone carbonyl, followed by deprotection and oxidation of the aldehyde,

    generated the Pictet-Spengler cyclization30 precursor (40). Under identical conditions

    utilized by Hoshino in his synthesis of (±)-lycoramine,31 the amide was treated with a

    formaldehyde equivalent to form the N-acyliminium ion, which was quenched by the

    adjacent aromatic ring to form the benzoazepine ring system (41). Reduction of the

    amide with LiAlH4 yielded (±)-galanthamine in 14 steps from commercially available

    starting materials.

    Research efforts have also revealed galanthamine to be an acetylcholinesterase

    inhibitor which enhances cholinergic function. In patients diagnosed with Alzheimer’s

    disease, (–)-galanthamine was well tolerated and produced significant improvement in

    attention and performance, leading to its FDA approval in 2001 under the RazadyneTM

    trade name.32 Sales of Alzheimer’s disease treatments, such as RazadyneTM, were in

    excess of 3.8 billion dollars (US) in fiscal year 2005, illustrating the demand for large

    quantities of (-)-galanthamine.33

    RazadyneTM is distributed in the United States by Ortho-McNeil Pharmaceuticals,

    a Johnson & Johnson subsidiary. Sano Chemia AG of Hungary claims to be the

    exclusive manufacturer and supplier of (-)-galanthamine to Johnson & Johnson. The

    Sano Chemia synthetic approach to galanthamine borrowed heavily from the previous

    work of Carroll and co-workers,21e which was based on the phenolic oxidation strategy

    pioneered by Barton and Kirby. Although the Sano Chemia route was highly-optimized,

    it still required 13 steps in a nine pot process, with a calculated overall yield of 13%

  • (Scheme 1.11).34 Clearly, there is room to improve on the synthesis of (-)-galanthamine

    (3) with regards to an effective and high-yielding process.

    O

    O

    O Br2

    MeOHreflux

    91% yield

    O

    O

    O

    Br

    95% H2SO4(aq)

    90 °C68% yield

    HO

    O

    O

    Br

    1) tyramine EtOH, reflux

    2) NaBH4 H2O, 5 °C

    92% yield

    HO

    O Br

    NH

    dioxane/DMF(30:1)reflux

    85% yield

    HO

    O Br

    N

    O

    K3[Fe(CN)6]K2CO3

    toluene/H2O(5:1)60 °C

    47% yield

    O

    O

    N

    O

    O

    HO OH

    p-TsOH (cat.)toluenereflux

    80% yield

    ON

    O

    O

    OO

    1) LiAlH4, THF, 60 °C

    2) 15% NaOH(aq), reflux3) 4M HCl(aq), 60 °C

    85% yield

    O

    O

    N

    O

    (-)-narwedine(1 mol %)

    EtOH/NEt3 (9:1)reflux → 40 °C

    75% yield

    O

    O

    N

    O

    OHHCO2Et

    HCO2H (cat.)

    OH

    1) L-selectride THF, -15 °C

    2) 48% HBr(aq), 10 °C98% yield

    ON

    O(-)-(3)·HBr

    HO

    H

    Br

    Br Br

    (42) (43) (44)

    (45)(46)

    (±)-(47) (±)-(48)

    (±)-(11) (-)-(11)

    Scheme 1.11. The Sano Chemia (-)-galanthamine·HBr (RazadyneTM) synthesis

    15

  • 16

    1.2 REFERENCES

    1) Cordell, G. A. Introduction to the Alkaloids: A Biogenetic Approach; John

    Wiley & Sons, Inc.: New York, 1981; pg. 533.

    2) Fernald, M. L. Gray’s Manual of Botany, 8th edition; American Book

    Company: NY, 1950; pg. 452.

    3) Raffauf, R. F. Plant Alkaloids: A Guide to Their Discovery and Distribution;

    Haworth Press, Inc.: Binghamton, NY, 1996; pg. 13.

    4) Harvey, A. L. The Pharmacology of Galanthamine and its Analogues.

    Pharmac. Ther. 1995, 68, 113.

    5) Pelletier, S. W. Chemistry of the Alkaloids; Van Nostrand Reinhold Company:

    New York, 1970; pg. 151.

    6) Hoshino, O. The Alkaloids: Chemistry and Biology; Cordell, G. A., editor;

    Academic Press: San Diego, CA, 1998; vol. 51, pg. 324.

    7) Zhong, J. Amaryllidaceae and Sceletium Alkaloids. Nat. Prod. Rep. 2005, 22,

    111.

    8) Dewick, P. M. Medicinal Natural Product: A Biosynthetic Approach; Second

    ed.; John Wiley & Sons, LTD: West Sussex, 2002.

    9) Wildman, W. C. The Alkaloids: Chemistry and Physiology; Manske, R. H. F.,

    editor; Academic Press: New York, 1960; vol. 6, pg. 289.

    10) Rinner, U.; Hudlicky, T. Synthesis of Amaryllidaceae Constituents – An

    Update. Synlett 2005, 3, 365.

    11) Proskurnina, N. F.; Yakovleva, A. P. Alkaloids of Galathus woronowi. II.

    Isolation of a New Alkaloid. Zhur. Obshchei. Khim. 1952, 22, 1899.

  • 17

    12) Barton, D. H. R.; Kirby, G. W. The Synthesis of Galanthamine. Proc. Chem.

    Soc. 1960, 392.

    13) Marco-Contelles, J.; Carreiras, M. C.; Rodriguez, C.; Villarroya, M.; Garcia,

    A. G. Synthesis and Pharmacology of Galanthamine. Chem. Rev. 2006, 106,

    116.

    14) Barton, D. H. R.; Kirby, G. W.; Taylor, J. B.; Thomas, G. M. Phenol

    Oxidation and Biosynthesis. Part VI. The Biogenesis of Amaryllidaceae

    Alkaloids. J. Chem. Soc. 1963, 4545.

    15) Eichhorn, J.; Takada, T.; Kita, Y.; Zenk, M. H. Biosynthesis of the

    Amaryllidaceae Alkaloid Galanthamine. Phytochem. 1998, 49, 1037.

    16) Barton, D. H. R.; Kirby, G. W. Phenol Oxidation and Biosynthesis. Part V.

    The Synthesis of Galanthamine. J. Chem. Soc. 1962, 806.

    17) Shieh, W.-C.; Carlson, J. A. Asymmetric Transformation of Either

    Enantiomer of Narwedine via Total Spontaneous Resolution Process, a

    Concise Solution to the Synthesis of (-)-Galanthamine. J. Org. Chem. 1994,

    59, 5463.

    18) Chaplin, D. A.; Johnson, N. B.; Paul, J. M.; Potter, G. A. Dynamic

    Diastereomeric Salt Resolution of Narwedine and its Transformation to

    (-)-Galanthamine. Tetrahedron Lett. 1998, 39, 6777.

    19) Heck, R. F. Palladium-Catalyzed Reactions of Organic Halides with Olefins.

    Acc. Chem. Res. 1979, 12, 146.

    20) Kametani, T.; Yamaki, K.; Yagi, H.; Fukumoto, K. Modified Total Synthesis

    of (±)-Galanthamine Through Phenol Oxidation. Chem. Comm. 1969, 425.

    21) A) Kametani, T.; Yamaki, K.; Yagi, H.; Fukumoto, K. Studies on the

    Syntheses of Heterocyclic Compounds. Part CCCXV. Modified Total

  • 18

    Synthesis of (±)-Galanthamine through Phenol Oxidation. J. Chem. Soc. (C),

    1969, 2602. B) Kametani, T.; Seino, C.; Yamaki, K.; Shibuya, S.; Fukumoto,

    K.; Kigasawa, K.; Satoh, F.; Hiiragi, M.; Hayasaka, T. Studies on the

    Syntheses of Heterocyclic Compounds. Part CCCLXXXVI. Alternative Total

    Syntheses of Galanthamine and N-Benzylgalanthamine Iodide. J. Chem. Soc.

    (C), 1971, 1043. C) Kametani, K.; Shishido, K.; Hayashi, E.; Seino, C.;

    Kohno, T.; Shibuya, S.; Fukumoto, K. Studies on the Syntheses of

    Heterocyclic Compounds. CCCXCVI. An Alternative Total Synthesis of

    (±)-Galanthamine. J. Org. Chem. 1971, 36, 1295. D) Holton, R. A.; Sibi, M.

    P.; Murphy, W. S. Palladium-Mediated Biomimetic Synthesis of Narwedine.

    J. Am. Chem. Soc. 1988, 110, 314. E) Szewczyk, J.; Lewin, A. H.; Carroll, F.

    I. An Improved Synthesis of Galanthamine. J. Heterocycl. Chem. 1988, 25,

    1809. F) Vlahov, R.; Krikorian, D.; Spassov, G.; Chinova, M.; Vlahov, I.;

    Parushev, S.; Snatzke, G.; Ernst, L.; Kieslich, K.; Abraham, W.-R.; Sheldrick,

    W. S. Synthesis of Galanthamine and Related Alkaloids – New Approaches. I.

    Tetrahedron 1989, 45, 3329. G) Szewczyk, J.; Wilson, J. W.; Lewin, A. H.;

    Carroll, F. I. Facile Synthesis of (±)-, (+)-, and (-)-Galanthamine. J.

    Heterocycl. Chem. 1995, 32, 195. H) Chaplin, D. A.; Fraser, N.; Tiffin, P. D.

    A Concise, Scaleable Synthesis of Narwedine. Tetrahedron Lett. 1997, 38,

    7931. I) Kita, Y.; Arisawa, M.; Gyoten, M.; Nakajima, M.; Hamada, R.;

    Tohma, H.; Takada, T. Oxidative Intramolecular Phenolic Coupling Reaction

    Induced by a Hypervalent Iodine(III) Reagent: Leading to Galanthamine-Type

    Amaryllidaceae Alkaloids. J. Org. Chem. 1998, 63, 6625. J) Krikorian, D.;

    Tarpanov, V.; Parushev, S.; Mechkarova, P. New Achievements in the Field

    of Intramolecular Phenolic Coupling Reactions, Using Hypervalent (III)

  • 19

    Iodine Reagent: Synthesis of Galanthamine. Synth. Commun. 2000, 30, 2833.

    K) Node, M.; Kodama, S.; Hamashima, Y.; Baba, T.; Hamamichi, N.;

    Nishide, K. An Efficient Synthesis of (±)-Narwedine and (±)-Galanthamine

    by an Improved Phenolic Oxidative Coupling. Angew. Chem. Int. Ed. 2001,

    40, 3060.

    22) A) Shimizu, K.; Tomioka, K.; Yamada, S.; Koga, K. A Biogenetic-Type

    Asymmetric Synthesis of Optically Active Amaryllidaceae Alkaloids: (+)- and

    (-)-Galanthamine from L-Tyrosine. Heterocycles 1977, 8, 277. B) Kodama,

    S.; Hamashima, Y.; Nishide, K.; Node, M. Total Synthesis of

    (-)-Galanthamine by Remote Asymmetric Induction. Angew. Chem., Int. Ed.

    2004, 43, 2659.

    23) Pilger, C.; Westermann, B.; Florke, U.; Fels, G. A New Stereoselective

    Approach Towards the Galanthamine Ring System via an Intramolecular

    Heck Reaction. Synlett 2000, 1163.

    24) Parsons, P. J.; Charles, M. D.; Harvey, D. M.; Sumoreeah, L. R.; Shell, A.;

    Spoors, G.; Gill, A. L.; Smith, S. A General Approach to the Galanthamine

    Ring System. Tetrahedron Lett. 2001, 42, 2209.

    25) Guillou, C.; Beunard, J.-L.; Gras, E.; Thal, C. An Efficient Total Synthesis of

    (±)-Galanthamine. Angew. Chem. Int. Ed. 2001, 40, 4745.

    26) Trost, B. M.; Tang, W. An Efficient Enantioselective Synthesis of

    (-)-Galanthamine. Angew. Chem. Int. Ed. 2002, 41, 2795.

    27) Trost, B. M.; Toste, F. D. Asymmetric O- and C-Alkylation of Phenols. J. Am.

    Chem. Soc. 1998, 120, 815.

    28) Hu, X.-D.; Tu, Y. Q.; Zhang, E.; Gao, S.; Wang, S.; Wang, A.; Fan, C.-A.;

    Wang, M. Total Synthesis of (±)-Galanthamine. Org. Lett. 2006, 8, 1823.

  • 20

    29) Fan, C.-A.; Tu, Y.-Q.; Song, Z.-L.; Zhang, E.; Shi, L.; Wang, M.; Wang, B.;

    Zhang, S.-Y. An Efficient Total Synthesis of (±)-Lycoramine. Org. Lett. 2004,

    6, 4691.

    30) A) Pictet, A.; Spengler, T. Über die Bildung von Isochinolin-derivaten durch

    Einwirkung von Methylal auf Pheny-äthylamine, Phenyl-alanin und Tyrosin.

    Chem. Ber. 1911, 44, 2030. B) Cox, E. D.; Cook, J. M. The Pictet-Spengler

    Reaction: A New Direction for an Old Reaction. Chem. Rev. 1995, 95, 1797.

    31) Ishizaki, M.; Ozaki, K.; Kanematsu, A.; Isoda, T.; Hoshino, O. Synthetic

    Approaches Toward Spiro[2,3-dihydro-4H-1-benzopyran-4,1'-cyclohexan]-

    2-one Derivatives via Radical Reactions: Total Synthesis of (±)-Lycoramine.

    J. Org. Chem. 1993, 58, 3877.

    32) Lilienfeld, S. Galanthamine – A Novel Cholinergic Drug with a Unique Dual

    Mode of Action for the Treatment of Patients with Alzheimer’s Disease. CNS

    Drug Rev. 2002, 8, 159.

    33) Whalen, J. Britain Stirs Outcry by Weighing Benefits of Drugs Versus Price.

    The Wall Street Journal, Nov. 22, 2005, pg. A1.

    34) Küenburg, B.; Czollner, L.; Fröhlich, J.; Jordis, U. Development of a Pilot

    Scale Process for the Anti-Alzheimer Drug (-)-Galanthamine Using

    Large-Scale Phenolic Oxidative Coupling and Crystallization-Induced Chiral

    Conversion. Org. Proc. Res. Dev. 1999, 3, 425.

  • Chapter 2: Studies Toward the Synthesis of (-)-Galanthamine

    2.0 DOUBLE CONDENSATION STRATEGY

    There have been several reported syntheses of galanthamine (1) and/or narwedine

    (2). Most of the syntheses are in excess of ten steps and almost all suffer from

    debilitating low yields at the key step (c.a. 40-55%).1 The low yields are often due to the

    use of phenolic or allylic oxidation chemistry, which results in overall yields of

    approximately 10%. In the syntheses of galanthamine and narwedine, the central issue

    has been the generation of a carbon-carbon bond, of which one carbon is a quaternary

    center, between the two six-membered rings of the carbon skeleton. A majority of

    previous syntheses have formed this bond via phenolic oxidation methods, with the

    remainder utilizing Heck-type couplings, as discussed in Chapter 1.1.

    O

    ON

    O

    O

    O

    OH

    N

    (-)-galanthamine (1) (-)-narwedine (2)

    O

    O

    OH

    N

    (-)-lycoramine (3)

    Figure 2.01. The structures of (-)-galanthamine, (-)-narwedine, and

    (-)-lycoramine

    My initial strategy, designed to enhance the efficiency of galanthamine synthesis

    and provide access to the Amaryllidaceae family, involved the generation of a substituted

    benzofuranone (4), which would be functionalized at the α-carbon and reduced to the

    21

  • lactol (5) (Scheme 2.01). A double aldol condensation could generate the enone ring

    system (6), followed by a Pictet-Spengler cyclization, as described by Martin and

    Garrison in their synthesis of (±)-lycoramine (3),2 to generate the benzoazepine ring

    structure and (±)-narwedine (2). The established resolution of (±)-narwedine into the

    desired single optical isomer3 made the racemate a sensible target intermediate in route to

    (-)-galanthamine.

    O

    O

    OO

    ON

    (±)-(2)

    O

    OH

    O

    O

    NR

    O

    O

    O

    (6)(5)(4)

    NR

    O

    CO2R

    Scheme 2.01. The double aldol condensation approach to (±)-narwedine

    Synthesis of the substituted benzofuranone (4) was envisioned as originating from

    the C-arylation of a dialkyl malonate. There are no examples reported in the literature in

    which 2-halo-6-methoxy-phenols successfully participate in the arylation of a malonate.

    Therefore, it was not surprising that 2-bromo-6-methoxy-phenol4 was unreactive under a

    variety of copper(I)-mediated arylation conditions which produced related

    benzofuranones,5 but yielded none of the desired benzofuranone (4) and only trace

    amounts of the mixed phenol/alkyl malonate esters. Additionally, palladium catalyzed

    conditions similar to those described by Hartwig,6 Buchwald,7 and Miura8 were also

    unsuccessful in forming the substituted benzofuranone. Arylation was achieved with

    2,6-dibromophenol9 (7) under copper(I) bromide conditions as reported by Konopelski,10

    to yield the 7-bromo-benzofuranone (8) (Scheme 2.02). Attempts to convert the

    22

  • aryl-bromide to an aryl-methoxy (9) under copper(I) chloride and sodium methoxide

    conditions11 yielded an intractable mixture of products.

    OH

    Br BrO

    O CO2Me

    O

    (9)

    OBr CO2Me

    O

    (8)

    NaH, CuBrMeCO2CH2CO2Me

    dioxane, reflux51% yield

    (7)

    CuCl

    NaOMe/MeOHDMF, reflux

    X

    Scheme 2.02. Formation of the benzofuranone core

    The multi-step conversion of o-vanillin into the 7-methoxy-3H-benzofuran-2-one

    was also considered,12 but was not pursued due to the lengthy synthesis required to

    generate the unstable benzofuranone core.13 Ultimately, efforts toward the

    benzofuranone/double condensation route were halted due to the inefficiency of the

    synthetic pathway. In the following subchapter, an alternative strategy is discussed.

    23

  • 2.1 BENZOAZEPINE STRATEGY

    The use of an intramolecular phenolate alkylation to generate a cross-conjugated

    cyclohexa-2,5-dienone and a quaternary carbon center was first described in 1957 by

    Winstein and Baird.14 Masamune utilized this method to establish the quaternary center

    of a divergent intermediate15 (13) in his 1964 syntheses of the diterpenes kaurene (15),16

    garryine,17 and atisine18 (Scheme 2.03). It was noted that only alkylbromide 12 cyclized

    to the dienone (13), while alkylbromide 11 did not react due to the steric repulsion of the

    axial C8 proton and the tetrahydropyranyl (THP) ether in the alkylation transition state.

    BnO

    CO2H(10)

    HO

    (11) OTHP

    Br

    HO

    OTHP

    Br

    (12)

    KOtBu

    tBuOHreflux

    OTHP

    O (13)

    (-)-kaurene (15)

    H

    H

    OH

    O (14)

    H3O+

    H

    H

    H

    H

    30% yieldfrom 10

    8

    8

    8+

    8

    6 steps

    Scheme 2.03. Masamune’s application of an intramolecular phenolate

    alkylation toward the synthesis of (-)-kaurene

    The most recent review of intramolecular phenolate alkylations describes the

    scope and limitations of this reaction.19 A majority of syntheses utilizing this method

    have been directed toward steroidal targets, but to our knowledge only one report has

    showcased this reaction’s utility in the synthesis of a non-terpenoid alkaloid.20

    24

  • The cross-conjugated dienone and quaternary carbon center of (±)-narwedine (2)

    maps well onto a strategy involving an intramolecular phenolate alkylation. Furthermore,

    an approach using this method would allow for an entirely new, and potentially efficient,

    synthetic entry into the Amaryllidaceae alkaloid family. With the overall goal of

    enhancing the efficiency of galanthamine (1) synthesis while providing access to the

    Amaryllidaceae alkaloid family, we envisioned a new approach to the target which

    originated from a functionalized biaryl compound (16) (Scheme 2.04).

    R1O

    O

    O

    O

    O

    OH

    NO

    ON

    O

    (±)-(2) (-)-(1)

    R1O

    ON

    O

    NR1O

    O

    OH

    N

    X

    (19)

    (18)(17)

    R1O

    O

    OH

    (16)

    R2

    Scheme 2.04. Benzoazepine approach to (-)-galanthamine

    Installation of the basic-nitrogen side-chain would introduce the tethered

    electrophile to the biaryl phenol (17). The electrophilic side-chain could exist in

    equilibrium between the open form and the aziridinium ion (18) via neighboring group

    participation of the nitrogen lone-pair21 and displacement of the primary leaving group.

    The phenol(ate) could cyclize onto the electrophilic carbon and/or the symmetrical

    aziridinium to generate the benzoazepine framework (19) of (±)-narwedine. Removal of

    25

  • the phenolic protecting group would allow for conjugate addition of the phenol to the

    dienone, yielding (±)-narwedine (2). As mentioned earlier, the established resolution of

    (±)-narwedine into the desired single optical isomer3 made the racemate a sensible target

    intermediate in route to (-)-galanthamine (1).

    Synthesis of the functionalized biaryl intermediate originated from the

    commercially available 2-bromo-3-hydroxy-4-methoxy-benzaldehyde (20), which was

    transformed to the methoxymethyl ether product (21) under conditions similar to those

    described by Fuchs and co-workers (Scheme 2.05).22 The protected aldehyde was

    coupled with a commercially available boronic acid (22), under Suzuki conditions,23 to

    yield the biaryl aldehyde (23). The aldehyde was subjected to reductive amination

    conditions with 2-(methylamino)-ethanol and NaBH(OAc)3 to install the basic-nitrogen

    side-chain (24).

    O

    MOMOO

    Br

    O

    MOMOO

    OH

    O

    HOO

    Br K2CO3MOMCl

    acetone50 °C

    99% yield

    Pd(PPh3)4 (2 mol %)

    2M Na2CO3(aq)EtOH/1,2-DME

    reflux73% yield

    (HO)2B OH

    O

    MOMON

    OH

    HOHN

    OH

    NaBH(OAc)3, AcOH1,2-DCE, reflux

    89% yield

    (20) (21)

    (22)

    (23) (24)

    Scheme 2.05. Construction of the functionalized biaryl intermediate

    26

  • The biaryl diol (24) contained all the required functionality to attempt a

    dehydration and cyclization to the cross-conjugated dienone and benzoazepine core

    structure (19). Initial attempts to utilize Dean-Stark dehydration conditions resulted in

    recovered starting material. Heating the diol (24) under refluxing conditions in the

    presence of a variety of othroformates resulted in the loss of the primary alcohol, and

    incorporation of the corresponding orthoformate alcohol to form a primary ether (25).

    This process is exemplified in Scheme 2.06 with triisopropyl orthoformate.

    O

    MOMON

    OH

    HO (iPrO)3CH

    reflux65% yield

    O

    MOMON

    OH

    O

    (24) (25)

    Scheme 2.06. Attempted dehydration and cyclization to the dienone

    The biaryl diol (24) was also subjected to Mitsunobu dehydration conditions.24 An

    interesting result occurred under the Mitsunobu conditions in which the cross-conjugated

    dienone (19) did not form, but instead, the parent biaryl aldehyde (23) was generated in

    high yield. This reaction pathway was rationalized through the mechanism shown in

    Scheme 2.07 in which the primary alcohol reacted with the activated phosphonium ion

    (26) to generate the primary leaving group. Then, the diethyl hydrazodicarboxylate anion

    revisited the molecule to remove the benzylic hydrogen (27), resulting in N-methylimine

    formation (28) and loss of the side-chain in the form of ethylene gas and

    triphenylphosphine oxide. It was assumed that the imine was hydrolyzed upon aqueous

    workup to produce the biaryl aldehyde (23). Raising or lowering the reaction

    27

  • temperature and altering the solvent concentration had no effect on the final product

    distribution.

    O

    MOMON

    OH

    HO

    O

    MOMO

    OH

    O

    DEAD, PPh3

    DCM, 0 °C98% yield

    (24) (23)

    O

    MOMON

    OH

    HO

    (26)

    O

    MOMON

    OH

    (27)O

    MOMON

    OH

    (28)

    NEtO2CHN

    Ph3P

    OEt

    O

    HNHCO2EtN

    EtO O

    -C2H4

    -Ph3P(O)

    H2O

    O PPh3

    Scheme 2.07. Attempted Mitsunobu dehydration of the biaryl diol

    Dialkyl azodicarboxylates have been reported as oxidants in the conversion of

    primary alcohols to the corresponding aldehydes,25 but when the biaryl diol (24) was

    subjected to diethyl azodicarboxylate, in the absence of triphenylphosphine, the diol was

    unreactive and only the starting material was recovered.

    Upon analysis of the results from the attempted dehydration of the biaryl diol

    (24), it was determined that a better leaving group than water might facilitate formation

    of the cross-conjugated dienone (19). Revisiting the biaryl aldehyde intermediate (23)

    allows for a change in the order of reactions to arrive at the new cyclization precursor. In

    the new scheme, the phenol of the biaryl aldehyde (23) was protected as a

    28

  • triisopropylsilyl ether (29), followed by reductive amination with 2-methylamino-ethanol

    and NaBH(OAc)3 to install the basic-nitrogen side-chain (30) (Scheme 2.08).

    O

    MOMON

    OTIPS

    HO

    O

    MOMOO

    OH

    TIPSClimidazole

    1,2-DCEreflux

    97% yieldO

    MOMOO

    OTIPS

    HN

    OH

    NaBH(OAc)3, AcOH1,2-DCE, reflux

    92% yield

    (23) (29) (30)

    Scheme 2.08. Synthesis of the silyl-protected biaryl alcohol

    Treatment of the primary alcohol (30) with methanesulfonyl chloride (MsCl) and

    iPr2NEt in dioxane at 23 °C resulted in a mixture of the primary mesylate (31) and the

    primary chloride (32). It was presumed that the primary mesylate was generated upon

    treatment with MsCl and a base, but the diisopropylethylamine hydrochloride salt did not

    appear to precipitate from the dioxane. Therefore, the chloride ion was still in solution,

    allowing it to displace the mesylate and form the primary chloride. The primary chloride

    was the exclusive product when the reaction mixture was heated at reflux under identical

    reaction conditions (Scheme 2.09).

    MsCliPr2NEt

    dioxanereflux

    91% yieldO

    MOMON

    OTIPS

    HO

    (30)O

    MOMON

    OTIPS

    MsO

    O

    MOMON

    OTIPS

    Cl

    (31) (32)

    Scheme 2.09. Synthesis of the biaryl primary chloride

    29

  • If the primary mesylate (31) was desired, the primary alcohol (30) was treated

    with MsCl and iPr2NEt in toluene at 23 °C, in which the diisopropylethylamine

    hydrochloride salt formed a visible precipitate, thereby removing the chloride ion from

    the equilibrium upon salt formation. The primary chloride (32) was selected as the

    intermediate of choice over the primary mesylate (31) due to the chloride’s superior

    stability during storage.

    The biaryl primary chloride (32) contained the requisite framework to attempt the

    intramolecular phenolate alkylation reaction to generate the cross-conjugated dienone

    (19). It was postulated that the removal of the silyl ether would result in phenolate

    formation, followed by an intramolecular alkylation. When the biaryl chloride (32) was

    treated with tetrabutylammonium fluoride (TBAF) in THF at 0 °C, or heated under

    reflux, the resulting product was the same new single spot by TLC. Attempts to isolate

    and characterize this product resulted in an intractable mixture with no carbonyl

    resonances upon IR analysis. A broad O–H stretch was seen in the IR, thus it was

    rationalized that the tetrabutylammonium fluoride reaction resulted in hydrodesilation

    and formation of the phenol (33), which decomposed upon dissolution on the rotary

    evaporator (Scheme 2.10).

    O

    MOMON

    OTIPS

    Cl

    O

    MOMON

    OH

    Cl

    (32) (33)

    TBAF

    THF

    Scheme 2.10. Attempted desilation and cyclization to the

    cross-conjugated dienone

    30

  • Treatment of the biaryl primary chloride (32) with silver(I) fluoride or AgBF4 in

    either THF, DMF, or CH3CN at 23 °C, or heating under reflux, also resulted in

    hydrodesilation (33). There was no evidence for the formation of the cross-conjugated

    dienone product (19) by IR or NMR analysis.

    The use of anhydrous powdered cesium fluoride with anhydrous DMSO or DMF

    at 23 °C resulted in the formation of a product other than the hydrodesilated product. The

    NMR and IR data indicated that the isolated compound did not contain resonances for the

    dienone protons or a carbonyl resonance, respectively. The spectral data, along with the

    mass spectrograph, correlated with the dimerization of the phenolate intermediate (34).

    X-ray crystallography confirmed the dimeric structure (36). It appeared that the use of

    cesium fluoride in DMSO or DMF formed the phenolate, but the intermediate reacted

    inter-, instead of intra-molecularly, to yield the dimer (Scheme 2.11).

    O

    OMOM

    OO

    MOMO

    O

    N

    N

    O

    MOMON

    OTIPS

    Cl

    O

    MOMON

    OMOMO

    ON

    O

    Cl

    (32) (34)

    (35)

    (36)69% yield

    X

    DMSO23 °C

    CsF

    Cs

    Scheme 2.11. Attempted cyclization to the cross-conjugated dienone and

    resultant dimerization

    31

  • 32

    Dimerization reactions simlar to that of the biaryl primary chloride (32) are

    precedented in reports of intramolecular phenolate alkylation substrates dimerizing,

    instead of forming the intramolecular phenolate alkylation product, presumably due to

    poor molecular overlap in the alkylation transition state.26 Attempts to avert the

    dimerization pathway by altering the solvent concentration and reaction temperature had

    no effect on the product distribution. As a result, it was postulated that the desired

    intramolecular alkylation to form the benzoazepine ring system was unlikely to occur

    under the described reaction conditions and an alternative intramolecular phenolate

    alkylation strategy was pursued.

  • 33

    2.2 PUMMERER STRATEGY

    In an attempt to facilitate the intramolecular phenolate alkylation strategy for the

    synthesis of (-)-galanthamine (1), new cyclization conditions were conceived. In the

    studies described in Chapter 2.1, the intramolecular phenolate alkylation substrate reacts

    inter-, instead of intra-molecularly. The exact reason(s) for this preference are unknown,

    but it is reasonable to assume that the undesired intermolecular alkylation pathway might

    have a more favorable molecular orbital overlap in the transition state than the desired

    intramolecular alkylation pathway, thereby giving rise to the dimeric product (36). Thus,

    it was hypothesized that changing the nature of the electrophile from an sp3 hybridized

    carbon to an sp2 hybridized carbon might enhance the molecular orbital overlap of the

    desired reaction pathway, and as a result, change the overall product distribution.

    To test the hypothesis of utilizing a sp2 hybridized electrophilic carbon, a strategy

    involving a Pummerer reaction27 was envisioned. The use of a functionalized biaryl

    intermediate (16), similar to that originally used in the benzoazepine strategy, would

    allow for a rapid exploration of the Pummerer strategy (Scheme 2.12). The cyclization

    substrate would originate from an amidosulfoxide (37), which could be activated to the

    electrophilic sulfonium ion (38). The sulfonium ion could be attacked by the

    para-carbon of the phenolic ether, followed by loss of the phenolic protecting group and

    formation of the cross-conjugated dienone core (39) of narwedine. The use of this

    strategy would also require the reductive cleavage of the phenyl sulfide and amide

    carbonyl groups after the dienone cyclization event to arrive at (±)-narwedine (2).

  • O

    R1O N

    OR3

    O

    (37)

    SO

    Ph

    O

    R1O N

    OR3

    O

    (38)

    SPhX

    R1O

    ON

    O

    (39)

    OSPh

    O

    ON

    O

    (±)-(2)

    R1O

    O

    OH

    (16)

    R2

    Scheme 2.12. Pummerer cyclization approach to (±)-narwedine

    Synthesis of the amidosulfoxide cyclization precursor (37) originated from the

    aforementioned biaryl aldehyde (23). Silyl ether formation with the biaryl phenol (40),

    followed by reductive amination of the pendant aldehyde with methyl amine, produced

    the biaryl secondary amine (41) (Scheme 2.13). The use of methanolic sodium

    borohydride and aqueous methyl amine suppressed appreciable formation of the benzyl

    alcohol byproduct encountered under Borch reductive amination conditions.28 Acylation

    of the amine with phenylsulfanyl-acetyl chloride,29 under basic conditions, yielded the

    amidosulfide (42). Oxidation of the sulfide with m-CPBA generated the amidosulfoxide

    (43) Pummerer cyclization precursor.

    34

  • O

    MOMONH

    OTBDMS

    O

    MOMOO

    OTBDMS

    NaBH4, MeOH23 °C

    73% yield

    40% H2NMe(aq)

    Cl

    OSPh

    iPr2NEt1,2-DCE

    23 °C78% yield

    O

    MOMON

    OTBDMS

    OSPh

    O

    MOMOO

    OH

    TBDMSClimidazole

    1,2-DCEreflux

    95% yield

    (23) (40) (41)

    (42)

    m-CPBA

    THF, 0 °C86% yield

    O

    MOMON

    OTBDMS

    OSO

    Ph

    (43)

    Scheme 2.13. Synthesis of the amidosulfoxide for exploration of the

    Pummerer cyclization conditions

    Treatment of the amidosulfoxide (43) with trifluoroacetic anhydride (TFAA) to

    generate the sulfonium ion, and thus allow for the cyclization to the cross-conjugated

    dienone, resulted in the formation of a vibrant yellow-green oil with spectral properties

    unlike those expected for the desired product. Upon further analysis, it was determined

    that the formation of the sulfonium ion had proceeded, but rather than be attacked by the

    para-carbon of the phenolic ether, it had been intercepted by the adjacent aromatic ring

    (44), resulting in the formation of a hydroisoquinolone skeleton (45) (Scheme 2.14). The

    hydroisoquinolone went on to lose a benzylic proton, resulting in the expulsion of the

    phenyl sulfide group and increased π-system conjugation of the isoquinolone structure

    (46). A similar result was reported by Yonemitsu and Oikawa, who utilized a Pummerer

    cyclization to synthesize functionalized 2-naphthols.30 The unstable isoquinolone (46)

    was the only identifiable component of a complex mixture of products from the

    Pummerer cyclization reaction. IR spectroscopy of the crude reaction mixture did not

    35

  • indicate the presence of a cross-conjugated dienone carbonyl resonance (c.a. 1660 cm-1).

    Raising or lowering the reaction temperature and altering the solvent concentration had

    no effect on the product distribution.

    O

    HON

    OTBDMS

    TFAA

    1,2-DCE23 °C

    53% yieldOO

    MOMON

    OTBDMS

    OSO

    Ph

    (43)(46)

    O

    RON

    OTBDMS

    OS

    Ph

    O2CCF3

    (44)R = MOM, H

    O

    RON

    OTBDMS

    OPhS(45)

    R = MOM, H

    H

    H O2CCF3

    Scheme 2.14. Attempted Pummerer cyclization to generate the core

    structure of narwedine

    Identical results were also observed the amidosulfide (42) was subjected to Lewis

    acid promoted Pummerer cyclization conditions.31 Under these conditions, the

    amidosulfide was treated with N-chlorosuccinimide (NCS) to form the α-chlorosulfide

    intermediate (47), which was reacted in situ with SnCl4 to promote formation of the

    sulfonium ion (Scheme 2.15). The sulfonium ion intermediate was trapped in the same

    manner as in Scheme 2.14 to produce the highly fluorescent and unstable isoquinolone

    product (46).

    36

  • O

    HON

    OTBDMS

    OO

    MOMON

    OTBDMS

    OSPh

    (42)(46)

    O

    MOMON

    OTBDMS

    OSPh

    (47)

    NCS

    PhCl0 °C

    SnCl4

    Cl

    52% yield

    Scheme 2.15. Lewis acid promoted Pummerer cyclization conditions

    After exploring the sp2 hybridized electrophilic carbon center (Pummerer)

    strategy and arriving at an interesting, yet undesired, result, a new strategy involving the

    use of an intramolecular phenolate alkylation was devised. In the following subchapter,

    the exploration of an alternative strategy is discussed.

    37

  • 38

    2.3 ETHER ALKYLATION STRATEGY

    In an attempt to avert the dimerization discussed in Chapter 2.1 (36) and the

    cyclization to the isoquinolone addressed in Chapter 2.2 (46), an alternative cyclization

    strategy was devised in which the phenolate would cyclize onto an electrophile to

    generate a six-membered ring intermediate. In the previous intramolecular phenolate

    alkylation strategies, the electrophile had been tethered at the benzylic position of the

    highly-substituted aromatic ring (17, 38). In the updated strategy, the electrophile was

    appended to the phenol, which was previously protected as a methoxymethyl ether. It

    was postulated that tethering the electrophile to the oxygen would not allow it to cyclize

    onto the adjacent positions because they were already substituted, thereby promoting

    cyclization onto the para-carbon of the phenolate. Additionally, removal of the

    methoxymethyl ether protecting group from the synthesis could facilitate a shorter route

    to (±)-narwedine (2) due to the deletion of steps required for the installation and removal

    of the methoxymethyl ether.

    Thus, the ether alkylation strategy was envisioned as originating from a similar

    biaryl intermediate (48) similar to the one utilized in the previously discussed

    approaches. Installation of the ether side-chain, which contained a two-carbon tether

    terminated by an electrophile, would generate the intramolecular phenolate alkylation

    precursor (49) (Scheme 2.16). Removal of the phenolic protecting group on the distal

    aromatic ring would form the phenolate (50), allowing it to cyclize onto the electrophile

    to form the cross-conjugated dienone (51) and the full carbon framework of galanthamine

    (1) and narwedine (3). Cleavage of the acetal (Y = OR) should allow the molecule to

    form the tricyclic dialdehyde (52). Reaction of the dialdehyde with methyl amine under

    double reductive amination conditions would yield the final ring required for the

  • synthesis of (±)-narwedine (2). As with the previous strategies, the established resolution

    of (±)-narwedine into the desired single optical isomer3 made the racemate a sensible

    target intermediate in route to (-)-galanthamine (1).

    O

    OX

    Y

    O

    O

    O

    O

    O

    O

    YO

    ON

    O

    (±)-(2)

    O

    O

    OR

    O

    XY

    (49) (50)

    (51)

    O

    O

    O

    (52)O

    O

    O

    HO

    OR

    O

    (48)

    Scheme 2.16. Ether alkylation approach to (±)-narwedine

    Synthesis of the substituted biaryl intermediate (48) required a different protecting

    group scheme than the approach used with the previously discussed strategies. The

    revised biaryl synthesis originated from commercially available 4-bromophenol (53).

    Triisopropylsilyl ether formation (54), followed by halogen-metal exchange and

    quenching with freshly distilled triisopropyl borate, yielded the boronic acid (Scheme

    2.17). The boronic acid was a gummy white solid that was stable for about one week on

    the bench top. Toluene azeotrope of the boronic acid yielded a boroxine (55) (the

    boronic acid trimer), which existed as a free-flowing white solid that was stable on the

    bench top for multiple months. Thus, the boroxine was the preferred intermediate over

    the boronic acid.

    39

  • OH

    Br

    TIPSCl

    imidazole1,2-DCE

    23 °C99% yield

    OTIPS

    Br76% yield

    1) nBuLi THF, -78 °C

    2) (iPrO)3B THF, -78 °C

    OTIPS

    BOO

    BO

    B(53) (54)

    (55) OTIPSTIPSO

    Scheme 2.17. Synthesis of the silyl-protected boroxine

    After establishing a route to the silyl-protected boroxine (55), the focus turned to

    the Suzuki coupling reaction. The coupling reaction employed the commercially

    available 2-bromo-3-hydroxy-4-methoxy-benzaldehyde (20) and the silyl-protected

    boroxine (55), which was available in two steps from the commercially available phenol.

    Since multiple steps were required to synthesize the boroxine, Suzuki reaction conditions

    were explored in which the boroxine was the limiting reagent. These conditions are in

    contrast to the numerous reports of Suzuki coupling reactions in which the boronic acid

    component is used in excess and the aryl-halide is the limiting reagent. 23

    The Suzuki cross-coupling conditions which were successful in the formation of

    the MOM-protected biaryl (23) were employed in the coupling of the silyl-protected

    boroxine (55) and the commercially available bromo-aldehyde (20). The reaction

    conditions resulted in the formation of numerous products and a 38% yield of the desired

    biaryl aldehyde (56). Some of the isolated byproducts included the hydro-deboronation

    product (57), the hydro-dehalogenated aldehyde (58), and the oxidized boronic acid

    phenol product (59) (Scheme 2.18).32

    40

  • O

    HOO

    OTIPS

    O

    HOO

    BrOTIPS

    BOO

    BO

    BAr Ar

    (55) (56)

    (20)

    O

    HOO

    OTIPS

    (57) (58)

    +

    OTIPS

    (59)

    +

    OH

    +Pd(PPh3)4 (2 mol %)

    2M Na2CO3(aq)EtOH/1,2-DME

    reflux

    Scheme 2.18. Synthesis of the biaryl aldehyde and unwanted byproducts

    Although phenyl transfer from triphenylphosphine (PPh3) to the palladium

    catalyst, and ultimately incorporation into the biaryl cross-coupling product, has been

    documented,33 no phenyl transfer was detected under the conditions screened.

    Regardless, the highly-effective tricyclohexylphosphine (PCy3) ligand34 was substituted

    for PPh3 to avoid the possibility of the aforementioned phenyl transfer complication. As

    a result of the modification in phosphine ligand, the catalyst was also changed to the

    commercially available and bench top stable [Pd2(dba)3] in place of the less stable

    Pd(PPh3)4.35 Catalyst loadings of 1-2 mol % were feasible, but resulted in yields of 30%

    and 51%, respectively. Slightly higher catalyst loadings of 3-4% resulted in higher

    yields. Attempts to conduct the reaction in the absence of phosphine with Pd(OAc)2,

    similar to the conditions reported by Novak36 and Hirao,37 resulted in approximately 50%

    yield of the biaryl aldehyde (56). Nickel catalysis38 was also explored, but that only

    produced an 18% yield of the biaryl product. After screening multiple catalysts and

    ligands, [Pd2(dba)3] and PCy3 were deemed the optimal combination.

    A judicious choice of base in the reaction suppressed the formation of the hydro-

    deboronation product (57). A strong mineral base promoted the formation of the hydro-

    deboronated product, while too weak of a mineral base resulted in sluggish rates of biaryl

    aldehyde (56) formation. The rationale for this observation is illustrated in Scheme 2.19,

    41

  • in which the boronic acid (63) interacts with a base during a Suzuki coupling reaction to

    form the reactive boronate ion (64). When the boronate ion is too reactive (i.e. as a result

    of a strong base), the formation of the hydro-deboronation product (65) predominates

    over the biaryl coupling pathway (68). If the base is too weak, then the formation of the

    activated boronate ion is sluggish (64), which in turn slows the overall rate of the biaryl

    coupling reaction (68). After screening multiple hydroxide, carbonate, and bicarbonate

    mineral bases, aqueous K2CO3 was selected as the optimal base for the cross-coupling

    reaction.

    LnPd(0)

    Pd(II)L

    LAr1 X

    Pd(II)L

    LAr1 Ar2

    Pd(II)Ar2

    LAr1 L

    Ar1 X

    Pd(II)L

    LAr1 O

    RH

    HO R

    X

    O

    R

    Pd(II)L

    LAr1 H

    Pd(II)H

    LAr1 L

    Ar1 H

    Ar2 B(OH)3B(OH)3

    Ar2 H

    H2O

    Ar2 B(OH)2HO

    Ar1 Ar2

    (60)

    (62)

    (61)

    (63)(64)

    (65)

    (66)

    (67)

    (68)

    (70)(69)

    (71)

    (72)

    (73)

    (74)

    Scheme 2.19. Generalized Suzuki biaryl coupling reaction pathways

    Exploration of the Suzuki biaryl coupling conditions also revealed that the

    removal of ethanol from the solvent system suppressed the formation of the hydro-

    dehalogenated aldehyde (58). It has been reported that Pd(II) species (62), which have

    42

  • 43

    undergone oxidative insertion with aryl halides (61), can go on to react with primary

    alcohols, in a hydride transfer process (70), to form palladium hydrides (72) (Scheme

    2.19).39 These metal hydrides can then undergo reductive elimination to expel the

    reduced arene (74) and regenerate the Pd(0) species (60).

    Thus, after considerable examination, a set of reaction conditions were established

    which consistently produced a 60-72% yield of the desired biaryl product (56), with the

    remainder of the material characterized as the oxidized boronic acid phenol byproduct

    (59). Similar oxidation byproduct distributions are also reported in the chemical

    literature where close inspection of the Suzuki biaryl coupling products have been

    conducted.40 In an attempt to solve this oxidation problem, the solvent was rigorously

    degassed by bubbling argon through the reaction mixture, prior to catalyst addition.

    Unfortunately, this method did little to absolutely suppress the oxidation byproduct

    profile. Thus it was determined that two options were available — subject the solvent

    system to a time consuming freeze-pump-thaw degassing regimen41 prior to running the

    Suzuki coupling reaction, or add an oxygen scavenger to the reaction mixture. The latter

    option was chosen due to its ease of operation upon scale-up. The oxygen scavenger,

    2,6-di-tert-butyl-4-methyl-phenol (BHT), was chosen due to its effective role as a

    stabilizing agent in laboratory-grade THF and its low cost. The use of 30-50 mol % of

    BHT suppressed any appreciable formation of the oxidation byproduct, allowing for use

    of the silyl-protected boroxine (55) as the limiting reagent in the Suzuki cross-coupling

    reaction with only 1.05 equivalents of the aryl-halide coupling partner (20). Under

    [Pd2(dba)3] catalyzed conditions, the two components were coupled to produce the biaryl

    aldehyde (56) in high yield (Scheme 2.20).

  • O

    HOO

    OTIPSO

    HOO

    Br

    [Pd2(dba)3] (4 mol %)

    P(Cy)3, BHT, K2CO3dioxane/H2O, reflux

    96% yield

    OTIPS

    BOO

    BO

    BAr Ar

    (55) (56)

    (20)

    Scheme 2.20. Optimized Suzuki biaryl cross-coupling conditions

    After optimization of the Suzuki cross-coupling reaction, conditions for the

    intramolecular phenolate alkylation were investigated. The biaryl aldehyde (56) was

    treated with chloroacetyl chloride and triethylamine to yield the biaryl α−chloro ester

    (75) (Scheme 2.21).

    O

    OO

    OH

    O

    Cl

    O

    OO

    OTIPS

    O

    Cl fluoride

    (75) (76)O

    HOO

    OTIPS

    (56)

    ClCl

    O

    NEt31,2-DCE

    23 °C92% yield

    heat

    Scheme 2.21. Attempt to generate the dienone intermediate from the

    biaryl α-chloro ester

    Treatment of the α-chloro ester product with TBAF in refluxing dioxane, or

    cesium fluoride in refluxing DMSO, yielded only the hydrodesilated product (76) with no

    evidence of the intramolecular phenolate alkylation product (51). Incorporation of

    sodium iodide or tetrabutylammonium iodide into the reaction to promote a Finkelstein

    44

  • process42 allowed for halogen exchange, but it did not facilitate the desired cyclization

    reaction. Likewise, treatment of the hydrodesilated product (76) with sodium hydride or

    potassium tert-butoxide yielded no cyclization product.

    The inability of the phenolate to cyclize onto the α-chloro ester under the

    described reaction conditions might be a result of the increased rigidity of the two-carbon

    tether due to the presence of the ester carbonyl, as noted in the studies of Masamune.16 As

    a modification of this approach, the α−halo acetal, as opposed to the ester, was pursued.

    There is precedent for the cyclization of an enolate onto an α-bromo acetal to form a

    six-membered ring product, as demonstrated in Fuchs and co-workers’ efforts toward the

    synthesis of the diterpenoid (±)-bruceantin.43 Based on this precedent, 1,2-dibromo-1-

    ethoxy-ethane was generated in situ by dropwise addition of ethyl vinyl ether to a 0 °C

    solution of bromine in dichloromethane. A dichloromethane solution of the biaryl phenol

    (56) and iPr2NEt was added to the reaction pot, resulting in alkylation of the phenol and

    formation of the primary halide acetal (77) (Scheme 2.22).

    OEt

    Br2, iPr2NEtDCM, 0 °C99% yield O

    OO

    OTIPS

    EtO

    Br

    O

    HOO

    OTIPS

    (56) (77)

    Scheme 2.22. Installation of the α-bromo acetal

    Treatment of the biaryl α-bromo acetal (77) with cesium fluoride in refluxing

    dioxane, toluene, acetonitrile, or 1,2-DCE did not facilitate cyclization to the

    cross-conjugated dienone (51) and only resulted in hydrodesilation. It was presumed that

    45

  • these solvents failed to generate the dienone because they were unable to reach

    temperatures high enough to promote the cyclization event, thus DMSO was explored.

    When the biaryl α-bromo acetal was treated with cesium fluoride in anhydrous DMSO at

    130 °C, cyclization and formation of the cross-conjugated dienone (79) occurred to

    produce the quaternary center and carbon framework of narwedine (2) (Scheme 2.23).

    Although a 65% yield of the desired product was isolated, there was a considerable

    amount of dark baseline material observed during column chromatography. Reaction

    temperatures below 130 °C with the CsF/DMSO conditions resulted in hydrodesilation

    (78) and no evidence of cyclization to the cross-conjugated dienone.

    O

    OO

    OTIPS

    EtO

    Br CsF

    DMSO130 °C

    65% yield O

    OO

    EtOO

    (77) (79)O

    OO

    O

    EtO

    Br

    (78)

    Scheme 2.23. Cyclization to the cross-conjugated dienone

    Under the CsF/DMSO cyclization conditions, an aroma similar to dimethyl

    sulfide was detected. It was considered that the dimethyl sulfide evolution may be due to

    a Kornblum oxidation44 of the α-bromo acetal (77). If the primary halide was displaced

    by the DMSO oxygen (80), followed by the loss of dimethyl sulfide (81), it could give

    rise to an aldehyde byproduct (82) (Scheme 2.24). Although the hypothetical aldehyde

    intermediate was never isolated under the DMSO reaction conditions, trace amounts of

    the biaryl diphenol (83) were isolated, which could have resulted from the decomposition

    of the dialdehyde acetal (82).

    46

  • O

    OO

    OTIPS

    EtO

    Br CsF

    DMSO130 °C

    (77) O

    OO

    O

    EtO

    (80)

    Br

    OS

    O

    OO

    O

    EtO

    O

    (81)

    S

    H

    -Me2S

    O

    OO

    OH

    EtO

    O

    (82)

    OAr

    decomposition

    O

    HOO

    OH

    (83)

    CsCs

    Scheme 2.24. Potential Kornblum reaction with the α-bromo acetal

    The use of DMF in place of DMSO was explored in order to achieve a

    high-boiling polar solvent medium which could also facilitate the cyclization to the

    cross-conjugated dienone without the potential to participate in the Kornblum oxidation

    side reaction. Since the use of wet DMF was unsuccessful in the aforementioned attempt

    to generate the cross-conjugated dienone (79), anhydrous DMF was examined. The DMF

    was stored over activated 4 Å molecular sieves and anhydrous cesium fluoride was also

    used to mitigate the presence of water. The DMF was decanted from the sieves, but

    during the process, molecular sieve dust entered the reaction. When the reaction was

    conducted in the presence of the sieve dust, none of the cross-conjugated dienone product

    (79) was formed, but a tetracyclic pyranone racemate (84) of similar polarity to that of

    the dienone was generated (Scheme 2.25). After column chromatography, the pyranone

    was the only identifiable product isolated. The tetracyclic pyranone racemate was

    47

  • isolated as an oil, which later crystallized into small plates, allowing for confirmation of

    the structure by X-ray analysis.

    O

    OO

    OTIPS

    EtO

    Br CsF

    4Å molecular sieve dustDMF

    130 °C24% yield(77) (84)

    OO

    EtO

    O

    O

    HH

    Scheme 2.25. Cyclization conditions to form the dienone in the presence

    of 4 Å molecular sieve dust, resulting in the generation of the pyranone

    Since the pyranone byproduct was observed when the reaction was conducted in

    the presence of the molecular sieve dust, it was hypothesized that the sieves and their

    contents played a role in the byproduct formation. Molecular sieves are known to be

    mildly acidic,45 and the sieve dust within the reaction mixture would contain some water

    that was absorbed from the DMF during the drying process. Thus, it was proposed that

    the tetracyclic pyranone (84) potentially arose from an acid-catalyzed rearrangement of

    the cross-conjugated dienone intermediate (79) (Scheme 2.26). The formation of the

    tetracyclic pyranone could result from the hydration of the cross-conjugated dienone (79),

    followed by a retro-aldol reaction to yield the ring-opened product (86). The ring-opened

    product could then participate in a cyclization reaction to yield the tetracyclic carbon

    skeleton (87). A 1,3-proton shift can generate the lactone (88), followed by a

    dehydration and another 1,3-proton shift to bring the styrene olefin into conjugation with

    the enone, resulting in the full conjugation of the α-pyranone olefins with the aromatic

    ring (84). It is hypothesized that the α-pyranone forms in preference to the γ-pyranone

    due to the increased aromaticity of the α-pyranones over the γ-pyranones.46 Additionally, 48

  • the formation of the α-pyranone allows for the full conjugation of the olefins within the

    molecule, whereas the γ-pyranone would isolate one of the olefins from the π-system

    unless the γ-pyranone were to exist in its higher energy pyrylium salt isomer.

    O

    OO

    OTIPS

    EtO

    Br CsF

    sieve dustDMF

    130 °C(77)

    (84)

    OO

    EtO

    O

    O

    HH

    O

    OO

    EtO

    (79) O

    O

    EtO

    O

    OH

    OHO

    O

    O

    EtOH

    O OH

    OH

    O

    O

    EtOH

    O OH

    OH

    (85) (86)

    O

    O

    EtOH

    O O

    OH

    (87) (88)

    1,3-protonshift

    -H2OH

    H

    Scheme 2.26. Postulated pathway for the formation of the tetracyclic

    pyranone racemate in the presence of 4 Å molecular sieve dust

    Anhydrous DMF, which had been dried over activated 4 Å molecular sieves, but

    allowed to settle and carefully decanted to minimize the carry-over of sieve dust, was

    combined with anhydrous cesium fluoride at 130 °C to produce the cross-conjugated

    dienone (79) in an improved yield over that achieved with DMSO. Furthermore, the

    reaction proceeded with no sign of oxidation or rearrangement byproducts. Thus, it

    appeared that a temperature of approximately 130 °C and anhydrous conditions (in the

    absence of sieves and sieve dust) were required for a successful intramolecular phenolate

    cyclization reaction with the biaryl α-bromo acetal (77) (Scheme 2.27). Drying the

    49

  • cesium fluoride under high vacuum, and a toluene azeotrope of the biaryl α-bromo acetal

    (77) prior to conducting the reaction resulted in a modest improvement in the yield.

    O

    OO

    OTIPS

    EtO

    Br CsF

    DMF130 °C

    90% yield O

    OO

    EtOO

    (77) (79)

    Scheme 2.27. Improved cyclization conditions to form the dienone

    The successful implementation of the intramolecular phenolate alkylation strategy

    avoided the low yielding phenolic oxidation reaction used previously to generate similar

    intermediates. The cyclization to the cross-conjugated dienone (79) formed the

    quaternary carbon center and the dienone ring required for the synthesis of narwedine (2)

    and galanthamine (1). The reductive amination of the aromatic aldehyde and latent

    aliphatic aldehyde to arrive at (±)-narwedine and (-)-galanthamine is discussed in the

    following subchapter.

    50

  • 2.4 AMINATION OF THE CROSS-CONJUGATED DIENONE

    The successful intramolecular phenolate alkylation to generate the

    cross-conjugated dienone, discussed in Chapter 2.3, formed the carbocyclic framework of

    (-)-galanthamine (1). To complete the synthesis, incorportation of methylamine was

    required at the carbons bearing the aromatic aldehyde and the ethyl acetal. Since both

    carbons were at the aldehyde oxidation state (89), it was envisioned that a reductive

    amination would allow for introduction of the methyl amine and subsequent formation of

    the narwedine (2) and galanthamine benzoazepine skeletons (Scheme 2.28).

    O

    O

    O

    O

    HOO

    ON

    O

    (±)-(2)(89)

    O

    O

    O

    (52)O

    OO

    OO

    EtOO

    (79)

    Scheme 2.28 Postulated reductive amination of the cross-conjugated

    dienone to arrive at narwedine

    A stepwise approach toward the amination was initially explored in an effort to

    closely monitor and understand each step of the process. Borch reductive amination

    conditions28 were considered to install the methylamine. The stability of the

    cross-conjugated dienone toward ethanol and NaBH3CN was explored since it contained

    multiple carbons at the aldehyde/ketone oxidation state which carry the risk of

    over-reduction with promiscuous reducing agents. After stirring the cross-conjugated

    dienone (79) with NaBH3CN in ethanol for one hour at 23 °C, there was no reaction by

    51

  • TLC. Upon heating the reaction at reflux for an additional hour, none of the carbonyls

    had been reduced, but the ethyl acetal of the aldehyde had formed (90) (Scheme 2.29).

    O

    OO

    EtOO

    O

    O

    O

    OEtO

    (79)(90)

    NaBH3CN

    OEtEtOHreflux

    68% yield

    Scheme 2.29. Exposure of the cross-conjugated dienone to NaBH3CN

    Since ethanol, in combination with the Borch reducing agent, generated the

    tetracyclic diethyl acetal product (90), the methylamine was installed via the application

    of NaBH(OAc)3 reductive amination conditions.47 The reaction resulted in the formation

    of a tetracyclic tertiary amine racemate (91) which existed as a mixture of diastereomers

    in an approximate ratio of 1:1, as determined by 1H NMR (Scheme 2.30). The acetal of

    the tetracyclic tertiary amine (91) was hydrolyzed to generate the lactol (92) in situ, but

    instead of remaining in the lactol form, or opening to the phenol and aldehyde, the lactol

    oxygen performed a conjugate addition onto the enone, resulting in the formation of the

    multicyclic tertiary amine acetal racemate (93) as a single diastereomer. The relative

    stereochemistry of the multicyclic tertiary amine acetal was determined by X-ray

    crystallography.

    52

  • O

    OO

    EtOO

    O

    O

    O

    N

    H2NMeNaBH(OAc)3

    THF23 °C

    EtO 3M HCl(aq)

    dioxanereflux

    53% yield(2 steps)

    NO

    O

    O

    O

    (79)(91)

    (93)O

    O

    O

    NHO

    (92)

    H

    Cl

    Scheme 2.30. Amination and hydrolysis to form the multicyclic tertiary

    amine intermediate

    The secondary amine multicyclic acetal (95) was formed in a similar manner as

    the tertiary amine multicyclic acetal (92). The cross-conjugated dienone (79) was treated

    with ammonium formate and NaBH(OAc)3 in refluxing ethanol to produce the tetracyclic

    secondary amine (94) as a mixture of diastereomers (Scheme 2.31).

    NHO

    EtO

    O

    OO

    OO

    OEtO

    NHO

    O

    O

    O

    HCO2NH4NaBH(OAc)3

    ethanolreflux

    94% yield

    3M HCl(aq)

    dioxanereflux

    84% yield(79)

    (94) (95)

    Scheme 2.31. Amination and hydrolysis to form the multicyclic

    secondary amine intermediate

    53

  • Ammonium formate was a suita