Towards the Synthesis of Aromatic Analogues of Himbacine, Potential Muscarinic Ligands

Home , Himbacine

TOWARDS THE SYNTHESIS OF AROMATIC ANALOGUES OF HIMBACINE, POTENTIAL MUSCARINIC LIGANDS

A thesis submitted in fulfilment of

the requirements for the degree of

Doctor of Philosophy

Xue Qin Shi

(B.E., M.Sc.)

Supervisor: Associate Professor Roger W. Read

SCHOOL OF CHEMISTRY

THE UNIVERSITY OF NEW SOUTH WALES MAY2000 CERTIFICATE OF ORIGNALITY

I hereby declare thatthis submission is my own workand that, to the best of my knowledge and belief, it contains no material previously published or written by another person nor material which to a substantial extent has been accepted for the award of any other degree or diploma of a university or other institute of higher learning, except where due acknowledgementis made in thetext .

CERTIFICATEOF ORIGINALITY

I hereby declare that this submission is my own worlc and to the best of my knowledge it contains no materials previously published or written by another person, nor material which to a substantial extent bas been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. AJJ.y contnbution made to the research by others, with whom I have worlced at UNSW or elsewhere. is explicitly acknowledged in the thesis.

I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged. 11

ACKNOLEDGEMENTS

I am very grateful to my supervisor Associate Professor Roger W. Read for his guidance, enthusiasm, encouragement and patience throughout the course of this study.

I also extend my thanks to my co-supervisor Prof. David StC. Black for his supervision while Roger was away on a Special Studies Program.

I am thankful to all technical staff in the School of Chemistry, especially to Dr.

Graha.111 Ball, Mrs. Hilda Stender and Dr. Jim Hook for their assistance in the 2-D n.m.r. experiments; to Dr. Joe Brophy and Mr. Rui Zhang for their useful help with the routine and high resolution mass spectra experiments.

I would like to thank the members of the Read group for their friendship and enjoyable discussions on many subjects. It is also grateful to Dr. Neil Donoghue for his kindly proof reading part of this thesis.

I am also grateful to the Department of Employment, Education, Training and

Youth affairs for the Australian Postgraduate Award (APA).

Last, but by no means least, I am deeply grateful to my family, my husband

Jiyuan and my lovely son Tian, for their love, constant understanding, continued support and patience during my study. ill

ABSTRACT

This thesis describes the development of a synthetic approach to aromatic analogues of himbacine, a potent muscarinc antagonist that is of interest to the study of age-related disease.

The studies have focused on strategies towards the lower (or southern) portion of the molecule and methods that involve the Michael reaction to assemble the carbon framework. An improved procedure for preparation of (+)-5-methyl-2(5H)-furanone as a Michael acceptor has been achieved in 51 % overall yield. Various methods of synthesis were investigated towards 2-(2-bromomethylphenyl)acetaldehyde as a

Michael donor. The molecule was finally synthesised from isochroman through hydrogen bromide cleavage and oxidation but it was unstable and not suitable as an intermediate. An alternative Michael donor, 2-(2-methoxymethylphenyl)acetaldehyde

N,N-dimethylhydrazone was eventually prepared and proved to be a more stable partner,, _j,,.- __ tMvirn dJ-r1 .-.d..,fwr11 f .,-, y/;,7 ~ and suitable intermediate for reaction. Conjugate additionnof' hydrazoner~Jtvith unsaturated ester was studied in model systems and the reaction applied to the above hydrazone and unsaturated lactone partner. A number of advanced intermediates with modified substituents were prepared but all proved difficult and did not undergo the ring closure desired for completion of the synthesis.

During the study, the stereochemical outcome of the Michael reaction was determined by spectroscopic methods including the analysis of unexpected products derived from alternative cyclisations. An unusual conformational exchange process was also discovered through n.m.r. analysis of one of the intermediates.

This work will provide a base on which future synthetic studies will be developed. iv

TABLE OF CONTENTS

CERTIFICATE OF ORIGINALITY

ACKNOWLEDGEMENTS 11

ABSTRACT 111

TABLE OF CONTENTS iv

CHAPTER 1. INTRODUCTION

1.1 Background 1

1.2 Muscarinic Receptor~M 2 Muscarinic Antagonists I

1.3 Alkaloid: Himbacine 4 6

1.4 Structure-Activity Relationships (SAR) 7

1.5 Structural Variants of Himbacine 11

1.6 Approaches to the Total Synthesis of Himbacine 4 Alkaloids 15

1. 7 Aim of the Study 22

CHAPTER 2. RESULTS AND DISCUSSION 23

2.1 Design ofHimbacine Analogue(s) 23

2.2 The Synthesis Plan 24

2.3 Preparation of Michael Acceptor, (+)-5-methyl-2(5H)-furanone 30 25

2.4 Preparation of Michael Donors 30

2.4.1 Attempted preparation of 2-(2-bromomethylphenyl)ethanol 79 31

Oxidation ofindene 78 31

via Claisen rearrangement ofally/ ether 101 35

2.4.2 Preparation of 2-(bromomethylphenyl)acetaldehyde 77 43 V

2.4.3 Preparation of 2-(2-methoxymethylphenyl)acetaldehyde N,N

dimethylhydrazone 117 47

2.4.4 Preparation of 2-(2-phenoxymethylphenyl)acetaldehyde 120

and its hydrazone 121 49

2.5 Michael Reactions 52

2.5.1 Reactions of enamine 131 with lactone 30 54

2.5.2 Reactions ofhydrazone derivatives 56

2.5.3 Reactions of arylmethyl bromide 141

with methyl crotonate/lactone 30 65

2.6 Preparation of Advanced Intermediates 70

2.6.1 Preparation of aldehyde 148 through solvolysis of

hydrazone 139 70

2.6.2 Preparation of acetal 149 76

2.6.3 Preparation of sulfide 153 80

2.6.4 Reactions of compounds 137 and 138

with hydrogen bromide 83

2.6.5 Preparation of aldehyde 158 90

2.7 Ring Closure 97

2.8 Summary and Future Directions 107

CHAPTER 3. EXPERIMENTAL 109

REFERENCES 173 CHAPTER 1. INTRODUCTION

1.1 Background

As the world population continues to grow and age, the problem of dementia will become even greater than it is today. 1 Alzheimer's disease (AD) is the most common form of dementia disorder. AD, a devastating neurodegenerative disorder that is characterized by dramatic personality changes, and global cognitive decline currently affects 15 million people worldwide,2 taking more than 100 000 lives each year. 3

Symptoms of age-related diseases are consistent with degeneration of the basal forebrain cholinergic system, i.e. the neural networks in the cortical and hippocampal regions of the brain. In these regions, neurotransmissions are mediated by acetylcholine whose release is regulated by binding to inhibitory M2 muscarinic receptors. Clinical studies of Alzheimer's disease and other age-related illnesses have been restricted largely to subjects that are in advanced stages of dementia, but show that while levels of muscarinic receptors generally remain high elsewhere in the central nervous system

4 (CNS), presynaptic M2 muscarinic receptors in the basal forebrain are depleted.

1.2 Muscarinic Receptors and M2 Muscarinic Antagonists

Muscarinic receptors are widely distributed in multiple organs and tissues and are critical to the maintenance of central and peripheral cholinergic neurotransmission.5

Molecular biological studies have demonstrated five distinct subtypes (ml-m5) of muscarinic acetylcholine receptors (mAChR) which share about 70% identity of amino acids in their seven trans-membrane segments.c,.io Each of these subtypes has been found to be localised in discrete areas of the brain or peripheral tissues. s-io Chapter 1 Introduction 2

Me'

1 pirenzepine 2 methoctramine 3 AF-DX 116

4 himbacine 5 4-DAMP 6

Chart 1. Compounds for characterization of muscarinic receptor

subtypes M1-~

Muscarinic receptor subtypes have also been divided into four individual groups:

M 1, M2, M 3 and M4, that have been characterized 11 by pharmacology using different antagonist affinities. For example, muscarinic M 1 receptors are characterised by a high affinity for pirenzepine 1 (Chart 1) and are found in high density in neuronal tissues and autonomic ganglia, 12 muscarinic M2 receptors have low affinity for pirenzepine 1 and high affinity for methoctramine 2, 13 AF-DX 116 3 14 and himbacine 4, 15 and are located mainly in cardiac tissue to modulate the release of acetylcholine, 16-21 whereas muscarinic Chapter 1 Introduction 3

M3 receptors display low affinity for pirenzepine 1 and high affinity for 4-DAMP 522

and p-fluorohexahydrosiladiphenidol 6,23 and are found in smooth muscles and glands.

At present, no selective antagonist is available for the muscarinic M4 receptor subtype.24

Also, the other cloned receptor (m5) is not well characterized owing to a lack of selective ligands. 25 There is some correspondence between the two subtype classifications, but the pharmacologically defined subtypes generally represent combinations of the gene cloned receptor subtype proteins.

Evidence has accumulated in recent years for the belief that selective muscarinic blockers might play an important role in the field of AD. For instance, it has been shown that M2 antagonists, such as AF-DX 116 3, enhance the release of acetylcholine in certain brain areas like cortex and hippocampus, both in vitro and in vivo. 26-28 This is due to a blockade of presynaptic receptors (marked as M2 sites in Figure 1) in the cortex

and hippocampus receptors which enhances release of acetylcholine (Ach).

The use of selective M2 antagonists could therefore be a new strategy to improve

memory and learning. According to this concept, a potential therapeutic agent would

have to possess good penetration through the blood brain barrier (BBB) and have a high

selectivity for M2 versus M 1 receptors, since the drug should not counteract its

presynaptic action by blocking postsynaptic M 1 receptors. 29 A schematic representation

of this working hypothesis is depicted in Figure 1. 29 Chapter I Introduction 4

postsynaptic

ACh

stimulation of this receptor results in IIMPULSE I presynaptic ~ inhibition of ACh release

Figure 1. Schematic representation ofmuscarinic receptors involved in cholinergic transmission in the forebrain

Muscarinic receptor antagonists that selectively blockade the inhibitory ~ 2 receptors, have therefore been sought as drugs for the treatment of cognitive disorders.

It has long been known that many muscarinic ligands require positive charge to bind to the receptor. This cationic requirement led to speculation that the binding site on the muscarinic receptor contains a negatively charged acidic residue.30 Relatively few compounds31 are effective in this regard, and those that are have widely differing chemical structures. They include dibenzodiazepinones,32 the most widely studied: AF

DX 116 3, AQ-RA 741 7, DIBA 8, BIBN 99 9; polymethylenetetramines:33 e.g. methoctramine 2, and mefurtamine; and alkaloids: imperialine34 and himbacine 4. 35 .14-.2 Chapter 1 Introduction 5

,,--/~-ClH 0 \._N:~u y (CH2b E(•y-N o Me~Me Me

7 X = N AO-RA 741 8 X=CH DIBA 9 BIBN 99

The Read group has begun a search for M2 muscarinic antagonists that might be used as diagnostic agents for the early onset of age-related diseases. In order to be successful, ligands must have high affinity for muscarinic receptors, have high selectivity towards the M2 subtype receptors, and be sufficiently lipophilic to permit blood-brain permeability. In this context a series of new amide analogues 10-13 of

DIBA 8, the most potent M2-selective antagonist described up to date has been synthesised.36•37 Analogues with groups that could be radiolabelled were of particular interest. The analogues demonstrated strong muscarinic binding affinity with 2-3 fold

M/M1 selectivity, but almost no M/M3 selectivity. One of the compounds, 12, actually

showed reasonable M3'M1 selectivity and efforts are continuing to understand the basis for the observed change in selectivity. Chapter 1 Introduction 6

10 H 11 C02Et 12 3-iodobenzoyl 13 4-fluorobenzoyl

In this study, a synthetic strategy towards modifications of an entirely different

class of muscarinic antagonist, namely the himbacine alkaloids, was investigated.

1.3 Alkaloid: Himbacine 4

Himbacine 4, an alkaloid from the bark of the Australian Galbulimima tree was first reported in 1956.38 The structural formula of himbacine 4 was reported in 1961,39 and stereochemical aspects of the structure were secured by X-ray crystallography in

1962.40 The alkaloid bears a characteristic tricyclic part, consisting of cis-fused y- lactone and trans-fused decalin moieties, connected to a trans-disubstituted piperidine through an (E)-double bond.40•41 There was relatively little interest in himbacine 4 until it was reported that it was a potent muscarinic antagonist that displayed selectivity for

42 5 the M2 receptor (Chart 2). -4 The molecule exhibits a ten- to twenty-fold higher affinity

15 for M2 muscarinic receptors than for M3 muscarinic receptors that mediate contraction.

Importantly, himbacine has demonstrated selectivity for presynaptic muscarinic receptors as compared with postsynaptic muscarinic receptors.42

Since blockage of presynaptic inhibitory muscarinic receptors leads to an elevation of synaptic levels of acetylcholine, it has been proposed that appropriate Chapter 1 Introduction 7

himbacine analogues might offset some of the losses in the cholinergic system that occurs in Alzheimer's disease.42

In contrast to other muscarinic receptor antagonists, such as gallamine, AF-DX

116 3 and methoctramine 2 that show cardioselectivity,46-48 himbacine interacts with muscarinic receptors in a competitive manner.49•50 Thus, as mentioned previously, it was considered a useful compound with which to study muscarinic receptors. However, it is not ideal in its blood-brain barrier permeability.

1.4 Structure-Activity Relationships (SAR)

Alkaloids related to himbacine 4, and certain derivatives (see Chart 2) are also

35 available and have been examined for their affinity for M2 and M3 muscarinic receptors, thus providing information about structure-activity relationships and structural features responsible for the selectivity. It should be noted that structures drawn in reference 35 for himbacine 4 and their derivatives were represented with the opposite stereochemistry at the stereogenic centres on the piperidine ring. This change in stereochemistry from structures in other papers was made without comment and might have arisen by error during publication.

Himbacine 4 showed a fifteen-fold selectivity for the M2 muscarinic receptor 4-2.. that was greater than that observed for any of the other compounds examined.»'Jt was the most potent antagonist when compared with himbeline 15, N-methylhimandravine

17 and himandravine 19, together with their dihydro-derivatives (Chart 2).

Reduction of the double bond linking the decalin ring system and the piperidine ring almost abolished selectivity in dihydrohimbacine 14. Removal of the N-methyl group in himbacine 4 to form himbeline 15 was also associated with a seven-fold decrease in selectivity. However the corresponding change in converting N- Chapter 1 Introduction 8

methylhimandravine 17 to himandravine 19 was not associated with any change in selectivity. These findings suggested that the N-methyl group was important for selectivity when the 2-methyl group on the piperidine ring was in the a., axial position

(himbacine 4 and himbeline 15) but not when it was in the ~. equatorial position (N- methylhimandravine 17 and himandravine 19). A double bond in the chain linking the piperidine and decalin ring systems also appeared important for high selectivity.

4 himbacine 14 dihydrohimbacine 15 himbeline 16 dihydrohimbeline

H 0 H 0 0

17 N-methyl- 18 N-methyldihydro- 19 himandravine 20 dihydro himandravine himandravine himandravine

Chart 2. Alkaloids related to himbacine 4

Muscarinic affinity was highest in the case of himbacine 4. Saturation of the double bond in the chain linking the two ring systems reduced affinity in the atria by three- to ten-fold for compounds with the N-methyl group on the piperidine ring Chapter 1 Introduction 9

(himbacine 4 and N-methylhimandravine 17), but there was no effect on affinity in the ileum. Alteration of the position of the 2-methyl group on the piperidine ring from a, axial, as in himbacine 4, to f3, equatorial as in N-methylhimandravine 17 was associated with a· 150- to 200-fold loss in affinity in atria. Thus high affinity in atria requires the methyl group in the 2-position on the piperidine ring to be in the a orientation. Also important, but having a lesser contribution was the presence of an N-methyl group and a double bond in the chain linking the ring systems.

Me .•' ii

4 himbacine 21 22

I iii

ii or iv

~ H OR 24R=C0Me 23 25 R = COC(Me)a

Reagents: i. LiAIH4; ii. Ac20, pyridine; iii. DIBAL; iv. Me3CCOCI, pyridine

Scheme 1 Chapter 1 Introduction 10

Modifications of the C-ring of the tricycle portion of himbacine have been examined (Scheme 1).45 Thus, the C-ring of himbacine's tricycle was opened by

complete reduction to the diol 21 through treatment with LiAlH4, and the diol 21 was converted into its diacetate 22.

Moreover partial reduction of himbacine 4 to lactol 23 with diisobutylaluminium hydride (DIBAL) enabled further treatment with acetic anhydride in pyridine to furnish the acetate 24, and with pivaloyl chloride in pyridine to generate the pivaloyl ester 25

(Scheme 1).

The carbonyl oxygen of himbacine 4 was also removed entirely to furnish the tetrahydrofuran 26.

The ring-opened diol 21 bound to M1 and M2 receptors with three- and twenty nine-fold less affinity, respectively, than himbacine 4, representing a nearly complete loss of Mi-selectivity. Meanwhile, the four compounds with intact tricyclic frameworks,

23-26, demonstrated decreased affinity at M1 and M2 sites when compared with himbacine 4. However, when compared with diol 21, the lactol 23 had approximately the same binding affinity at the M1 receptor, but only two and a half-fold better binding at the M2 receptor. It appeared then that an intact C-ring framework alone was not sufficient to ensure M2 selectivity and potency. Chapter 1 Introduction 11

Compound 26 differs from himbacine 4 only in that the carbonyl oxygen has been removed, thus tricycle 26 resembles himbacine sterically, but is different in its electrostatic field. The compound had much less affinity and poorer binding than himbacine 4 at the M2 receptor. The results indicated that the carbonyl oxygen of the C ring and a closed C-ring were necessary for high affinity and selectivity at the M2 receptor subtype.

1.5 Structural Variants of Himbacine

Efforts have been made to chemically modify the himbacine structure in order to further study its structure-activity profile and to improve potency and the relative selectivity for the presynaptic receptor.

A highly simplified himbacine analogue, 27, which embodies only its more hydrophilic elements, i.e. the y-lactone ring connected via an olefinic appendage to a piperidine ring was synthesised through a conjugative addition approach.43

The piperidine portion of the molecule was constructed from (±)-pipecolic acid

28 through a sequence involving an initial chemical resolution of the acid followed by reduction, tosylation, and iodide formation. Reaction of the (S)-iodide with sodium benzenesulfinate followed by detosylation and reductive methylation afforded the desired piperidinyl sulfone 29 (Scheme 2).

The optically pure lactone (+)-5-methyl-2(5H)-furanone 30, acquired from D glutamic acid 31 by a sequence of fairly well precedented steps,51 was alkylated with acetaldehyde N,N-dimethylhydrazone under the influence of Cu(I) catalysis. The hydrazone 32 was cleaved by ozone treatment to afford the aldehyde 33, and Julia coupling between sulfone 29 and aldehyde 33 in the standard manner2 afforded analogue 27 (Scheme 2). Chapter 1 Introduction 12

28 (:!:)-pipecolic acid 29

Me (Me)2N·N~ Me Qo ___ .::-0 0 0 31 30 32

j ii

O~HMe iii-vi .::- 0

0 0 27 33

Reagents: i. acetaldehydeN,N-dimethylhydrazone, cat. Cu(l)~i. 0 3, CH2Cl2; iii. n-Buli, THF; iv. 29; v. PhCOCI; vi. 6% Na-Hg, MeOH, THF.

Scheme2

The analogue 27 was 260-fold less potent than himbacine 4 at brainstem M2 receptors. Removal of the decalin ring also reduced selectivity for the M2 receptor; himbacine 4 was about twenty-fold selective for the M2 over the M1 receptor while analogue 27 was only eight-fold selective. Similarly, while himbacine 4 was about fourteen-fold selective for the M2 receptor, with respect to the M3 receptor, analogue 27 was only about four-fold selective in the same comparison.43

It has been proposed that the energetically costly conformational reorientation of

27, which may be required for binding to the M2 or M4 receptors, together with loss of the hydrophobic binding component of himbacine's decalin moiety, contribute to its Chapter 1 Introduction 13

poor binding.43 Thus, it is presumed that the decalin moiety plays an important role in conferring M2 selectivity to himbacine 4.

In another endeavour it was postulated that alternative tricyclic motifs could equally mimic himbacine's hydrophobic southern portion, and a 9,10-dihydroanthracene substituent was used as a simplified replacement for the himbacine tricyclic system

(Scheme 3).44 Thus, the previously described sulfone 29, in racemic and both enantiomerically pure forms 29a and 29b were subjected to modified Julia coupling with 9-anthraldehyde to generate alkenes 34, 34a, and 34b and the double bond isomerized products 35, 35a and 35b, respectively (Scheme 3). Under the coupling conditions the central ring of the tricycle was reduced to the dihydroanthracene ring system. Rearomatization of the central ring of 34a using dichlorodicyano-benzoquinone

(DDQ) in dichloromethane furnished anthracene 36.

i-iv + PhS02~ Me

29a (R) 34a (R) 35a (R) 29b (S) 34b (S) 35b (S) 29 (racemic) 34 (racemic) 35 (racemic)

Reagents: i. n-BuLi, -7a0c, THF; ii. 9-anthraldehyde; iii. BzCI; iv. 6% Na-Hg, NaHP04, MeOH, THF.

Scheme3

The intervening chai_n was also extended by one methylene unit to give 37, in

order to examine the effect of the tether flexibility on receptor selectivity. It was

postulated49 that the orientation of the nitrogen head group in relation to the tricycle was Chapter 1 Introduction 14

of major importance for Mi/M1 selectivity. Adjustment of the chain length would have

allowed more conformational mobility between the two components, but there was no

appreciable effect on binding of 37 relative to compound 34a.

36 37

The synthetic compounds 27, 34 and 35 were highly potent at both M1 and M2 binding sites. For the anthracenyl derivative 36, aromatization of the central ring, whether by affecting geometry, electronics, or both, caused a thirty-fold loss in potency at the M2 receptor and a four hundred-fold loss in potency at the M1 receptor. In contrast, dihydroanthracene 34a had an even higher affinity than himbacine 4 at the M2 receptor, and was almost as potent as DIBA,53 the most potent Mi-selective antagonist described to date. The assay results support the concept that the stereogenic centre on the piperidine ring might play a role in binding at the M2 subtype, since binding at the

M1 receptor was virtually identical among 34a, 34b and 34, while slightly different at the M2 receptor. In general, those analogs containing an £-double bond showed preferential binding to the M1 receptor, while those compounds 35a, 35b and 35 derived from the double bond rearrangement displayed non-selective binding and slightly lessened affinity when compared to their isomers. Chapter 1 Introduction 15

More recently, a parallel synthesis of racemic himbacine analogues, based on a previously described total synthesis of himbacine54-58 (see following section) was reported (Scheme 4).59

ONs O R N IBPO~ NvV' H I Me 3 steps ,Me •' ,Me ·' -- 0 OR ~ i or ii 0 0 0 T 38 39 40 (X)

Reagents: i. K2CO3, Kl, CH3CN, 130°C; ii. DMAP, CH3CN, r.t.

Scheme4

The synthesis proceeded from the trienyne ester 38 through a common advanced intermediate, nosylate 39, which was reacted with various commercially available heterocyclic secondary amines. Several of the simplified synthetic analogues, e.g. 40 and 41 showed potency comparable to that of himbacine 4, albeit lacking the desired selectivity.

1.6 Approaches to the Total Synthesis of Himbacine 4 Alkaloids

Accordingly, himbacine 4 has attracted much recent interest from medicinal and synthetic organic chemists. Despite knowledge of its structure and the obvious biological importance of the compound, there had been only one reported total synthesis of himbacine 4 at the outset of this project. It was synthesised for the first time by Chapter 1 Introduction 16

Hart's group in 1995.54 Since then, total syntheses of himbacine 4 have been achieved by Chackalmannil's group55•60 and Terashima's group.41

The first total synthesis of ( +)-himbacine 4 (Scheme 5)54 involved the intramolecular Diels-Alder reaction of unsaturated thiol ester 42 and proceeded through reduction of the resulting tricyclic thiol ester 43, conversion of the alcohol 44 in two steps into the phenylsulfide 45, transformation of the lactone into a cyclic acetal 46, and finally oxidation of the sulfide into its sulfone 47. Attachment of the piperidine ring portion 48 was then achieved through Julia coupling (Schemes 6 & 7).

COSEt ii ~:· H H 0 0 43 ... . ( 44 A = CH2OH 42 Ill, IV 45 A = CH2SPh

Iv, vi

Me ,•' vii

OMe

47 46

Reagents: i. SiO2-Et2AICI, PhMe, 40°C, 96 h; ii. Raney-Ni, EtOH, Et2O; iii. Jr TsCI, pyridine; iv. KO-t-Bu, DMSO, PhSH; v. i-Bu2AIH, Et2O, hexane; vi. BF3.Et2O, MeOH, CH2C12_; vii. TTrCPBA, CH2Cl2, NaHCO3.

Schemes Chapter 1 Introduction 17

i, ii iii x_}) x~-"Me I I OHCJ),,,Me I BOC Boe BOC

X=OH X=OTBS 48 X=OTBS X=OH

Reagents: i. s-Buli, Et2O, TMEDA; ii. Mel; iii. n-Pr4NRuO4, NMO, CH2Cl2.

Scheme6

i, ii, iii iv

R=BOC V t4915 R=H vi 4 R=Me

Reagents: i. n-Buli, DME; ii. 48; iii. 6°1> Na(Hg), MeOH, Na2HPO4; iv. H2Cr2(h, acetone; V. TFA, CH2Cl2: VI. NaBH3CN, HCHO

Scheme7

Since this was the first successful total synthesis of himbacine 4, it provided a protocol for conducting analogue studies, either via total synthesis or via intermediates.

From the standpoint of chemistry, the synthesis featured an intramolecular thioester

Diels-Alder reaction and suggested that this little-explored family of dienophiles might find some general use in synthesis. On the other hand, this total synthsis was relatively long and proceeded via a longest linear sequence of twenty steps.

A shorter, highly convergent approach to (+ )-himbacine, was communicated by

Chackalmannil' s group in 199655 and described in full in 199960 (Scheme 8). The Chapter 1 Introduction 18

approach involved only eleven linear steps, and construction of the alkaloid through an enantioselective, intramolecular Diels-Alder reaction of an appropriately functionalized substituted tetraene 51. The tetraene 51 was designed to bear the entire latent carbon framework of himbacine 4, including the trans-2,6-disubstituted piperidine ring system.

The methodology of this total synthesis of himbacine 4, especially the intramolecular Diels-Alder reaction that was used in the construction of the tricyclic ring system, could be further applied in the SAR studies of this series of compounds .

.,,Me .,,Me

ii-iv

Me OH HOOCV ' Me ~ 0 R 0

50 51 R=~-H R =a-H

vi, vii

15 R= H 4 R=Me

Reagents: i.1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride, 4-dimethylamino pyridine, 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO) (1%), CH2Cl2: ii. toluene, TEMPO, 1as 0 c; iii. DBU; iv. (Boc)2O, NaOH; v. Ra-Ni, H2, MeOH; vi. TFA, CH2Cl2; vii. HCHO, NaBH3CN, CHaCN.

Schemes Chapter 1 Introduction 19

More recently, a novel total synthesis of himbacine 4 was accomplished by

Terashima's group by using an exo-selective, intermolecular Diels-Alder reaction of the tetrahydroisobenzofuran 52 with the chiral lactone 30 as the key step, (Scheme 9).41

Thereafter, hydrogenation of the double bond in the cycloadduct 53 from the sterically less hindered convex face afforded the saturated compound 54, which was subjected to base catalysed ~-elimination of the oxygen bridge, followed by double bond isomerization and catalytic hydrogenation to give the tricyclic alcohol 55. Further modification over eight steps then provided the known sulfone 47, and therefore the route merged with the synthesis of Hari.i56 According to Hart's protocol, sulfone 47 was finally converted to himbacine 4 in 5 steps.

This approach was more convergent in construction of the tricyclic system than the previous two approaches. However, it required numerous linear steps to complete

the synthesis.

Me H 1',1e ~Me (Co + Qa Olto ii Y.-0:-(0 0 H 0 H H o

52 30 53 54

j iii

OH 8 steps djf iv, V H H o

47 55

Reagents: i. 5 M LiCIO4 in Et2O; ii. H2, 10% Pd-C, EtOH; iii. LiN(TMS)2, THF; iv. DBU, toluene; v. H2, PtO2, EtOH.

Scheme9 Chapter 1 Introduction 20

Two closely related but independent studies towards himgravine 60 were also 51., reported.58" 61 They involved an intramolecular Diels-Alder strategy resembling that of s-i,bl Hart (Scheme 5). In one, a thermal reaction of triene 56 provided an 8:5 mixture of isomeric tricyclic aldehydes 57 and 58 (Scheme 10). The desired endo adduct 57 was the major isomer and its correct relative configuration at the five contiguous stereogenic centres was anticipated from the endo-mode of addition of the diene portion to the least hindered face of the (E)-dienophile.62 Coupling of the piperidine ring portion to aldehyde 57 was achieved through Wadsworth-Homer-Emmons olefination rather than

Julia coupling, using a racemic iminophosphonate 59 to give the £-isomer 60 as a mixture of diastereomers (Scheme 10).

CHO Me H : H ~ . ~· + ., 0 ~A o A o

56 57 58

ii, iii ii, iv

P(OEt)2 '17II 0

59 60

Reagents: i. tolu~ne, 170°C (di-t-Bu-p-cresol); ii. LDA (1 eq), THF; iii. (EtO)2POCI, THF; iv. 57.

Scheme 10 Chapter 1 Introduction 21

In the other approach57 to himgravine 60 (Scheme 11), the precursor 56 was

cleverly converted in situ into a related oxycarbenium ion 61 that underwent

spontaneous Diels-Alder reaction to give a key tricyclic intermediate 62. The acetal 62 ~d;J, ~,:f"t was transformed into the corresponding alcohol 63,: ·. A I con~ersion into aldehyde woulJ. brio 4- '\ 57 · tlae 1wo syntheses to a common intermediate.

+/\ /7 Ao OTMS 0 0

56 61 62

j ii, iii

Reagents: _i_. TMSOCH2CH2OTMS, T~§,OTf, CH2Cla: II. TsOH, CH3COCH/H2O; Ill. Na8H4, EtoH. ~- H o 63

Scheme 11

It will be noted that all the literature synthetic approaches to himbacine alkaloids have Diels-Alder cycloaddition reactions to construct the major tricyclic part of the molecule. Substituents then reside in the desired configuration as a consequence of the mechanistic demands of the reaction. In most cases, the researchers have employed a probable biomimetic strategy involving an intramolecular cycloaddition to improve selectivity. While elegant in their design, these syntheses require considerable effort to prepare the necessary Diels-Alder precursors. Chapter I Introduction 22

1. 7 Aim of the Study

The aim of the present work was to develop a synthetic approach to aromatic variants of himbacine 4. Ultimately, it was planned to synthesise iodine and fluorine

substituted analogues that, if found to be suitably M2 selective, could be used in radiolabelled form as diagnostic agents in the study of age-related diseases, through use of tomographic emission techniques, like positron emission tomography (PET) and

single photon emission computed tomography (SPECT). Emphasis was placed on

assembly of the southern part of himbacine analogue 64.

One approach that was considered was to assemble the tricycle through a

cycloaddition reaction. The synthesis would resemble that in Scheme 9 and would

involve reaction between lactone 30 and an isobenzofuran. Closer examination revealed

that this strategy required lengthy linear sequences to construct the appropriate

isobenzofuran precursors, and it was considered unattractive.

An alternative that was pursued was to assemble the tricycle through a tandem

sequence involving conjugate addition of a substituted aromatic precursor on to lactone

30 followed by cyclisation to close ring B (Scheme 12). This sequence bears some

resemblance to syntheses of podophylotoxin lignans 100 and has precedent, at least in the

conjugate addition step, in the synthesis of simplified himbacine analogues through

Scheme 2. 23

CHAPTER 2. RESULTS AND DISCUSSION

2.1 Design of Himbacine Analogue(s)

In this study it was proposed to investigate himbacine analogues 64 in which ring A of himbacine 4 was replaced by an aromatic ring. The effect of this change alone was not considered during earlier structure-activity studies (section 1.4, 1.5),35,43-45,59,63 although it was deduced that having a ring A was necessary for activity.

Initial studies have focused on synthesis of the lower (or southern) portion of structure 64, and in order to simplify the chemistry, analogues without halogens or other substituents in the aromatic ring (X = H) were considered in the first instance.

Me ,Me .•' ·'

4 himbacine 64

2.2 The Synthesis Plan

The target molecule, 64, contained only four of the normally six stereogenic centres associated with the tricyclic portion of himbacine 4, and it appeared relatively easily accessible by the disconnection route outlined in Scheme 12. A logical disconnection occurs at the double bond that connects the piperidine ring to the tricyclic portion. The study by Hart54 indicated that Julia coupling of two such reactants proceeded best when the aldehyde partner was attached to the piperidine ring, e.g. as in Chapter 2 Results and Discussion 24

48. Synthetic efforts therefore were aimed at sulfone intermediates, such as 65, although flexibility in the types of groups through which the piperidine ring could be attached was sought. Further disconnection of sulfone 65 suggested the possible precursors, known lactone 30 and aromatic structure 66. A diastereoselective Michael addition reaction by an anion from synthetic block 66 on to lactone 30 should generate an advanced intermediate with the correct relative stereochemistry of groups attached to the lactone ring. Subsequent intramolecular substitution closure through direct displacement of the Z group or a condensation reaction should then afford a tricyclic unit that could be converted into sulfone 65. Q.,,Me N,Boc CHO .. ,,Me 48 9N,M + X . = S02Ph ~,Me A'-. J:H ~e x~b x CJJ!ko = H O H 0

64 65 66 30

Scheme 12

It was anticipated that this approach would provide access to compound 64 with an aromatic ring through which fluorine or iodine could be introduced, and with retention of the features that are known to be necessary for muscarinic potency and selectivity. Jn practice, the synthesis of tricycles like 65 proved more challenging than at first thought and this target molecule was not attained. However, much background Chapter 2 Results and Discussion 25

chemistry was investigated and an advanced intermediate prepared. The main reactions involved in the study, namely, Michael reaction, intramolecular cyclisation and the preparation of precursors and intermediates to each are discussed in tum below.

2.3 Preparation of Michael Acceptor, (+)-5-methyl-2(5H)-furanone 30

Compared with some of the other classes of conjugate acceptors, a,~ unsaturated lactones have received little attention.64 However, acce~s to the optically active lactone 30 was an important element in the planned asymmetric synthesis of tricycle 65. Lactone 30 is not commercially available but was employed for the conjugate addition approach to himbacine model substances (Scheme 2)43 and in the cycloaddition approach to himbacine 4 itself by Terashima (Scheme 9).41 Indeed, the biological importance oflactone 30 has prompted several synthetic approaches51 •65"67 to the molecule and it has been used as a synthetic intermediate on a number of occasions.

The most satisfactory synthesis65 of 30 in racemic form appears to be through a sequence involving opening of propylene oxide with the dilithio salt of thiophenoxyacetic acid, oxidation and thermal elimination (Scheme 13). This procedure is versatile because it provides various substituted ~a.11-butenolides by using different epoxides.

iv ...... I, II, Ill nSPh nSOPh ---Me O O v Me-¥0 Me O 0

Reagents: i. LDA, THF, o°C; ii. propylene oxide, THF, -60°C, r.t.; iii. H2SO4 (cat.), reflux; iv. sodium metaperiodate, MeOH, r.t.; v. toluene, 110°C

Scheme 13 Chapter 2 Results and Discussion 26

The substance has also been prepared in enantiomerically pure form from D glutamic acid. 51 This approach was used during the synthesis of model lactones in the study of himbacine 4 analogues (Scheme 2).43 It provided the compound in good yield and optical purity but the unavailability of D-glutamic acid in our laboratories and its expense, unlike the natural L-glutamic acid, made this an unattractive approach.

Another asymmetric synthesis was achieved from L-tartaric acid by a method requiring monosaponification of the acetonide diester 67 and a relatively linear chain extension from the half ester 68 through alcohol 69 and tosylate 70 as depicted in

Scheme 14.66 The low overall yield of lactone 30 from L-tartaric acid (only 10%) discouraged any attempts at this method.

.. 69 R= OH ; ,e_'::-C,i-faOH 11 ( 70 Al-= OTs; t:<'= CHl-Ors

vi V

Reagents: i. UAIH4, ether; ii. TsCI, Py; iii. MeOH, Lewatit S-100 resin; iv. NaCN, DMSO; v. HCI, sat. MeOH; vi. MsCI, Et3N, CH2Cl2

Scheme 14

An alternative asymmetric synthesis from ethyl (S)-(-)-lactate 71 (Scheme 15)67 was more attractive because of the availability and low cost of the starting materials, the Chapter 2 Results and Discussion 27 reported efficiency of the process (4 7% overall), and the seeming ability of the process to be scaled-up.

71 72 73

! iii EtO~O

V CMe iv C02Et 76a + 30 74 EtO~O

Me~C02Et 76b

Reagents: i. CH2=CHOEt, PPTS; ii. LiAIH4; iii. (COCl)2, DMSO, Et3N; iv. Ph3PCHC02Et 75, MeOH; v. MeOH-H20, H2S04 (cat.).

Scheme 1S

In practice, ethyl (S)-(-)-lactate 71 was protected with ethyl vinyl ether and the crude ester 72 reduced with lithium aluminum hydride according to the literature method.68 The alcohol 73 was obtained in 95% overall yield as a mixture of diastereomers due to the new acetal stereogenic centre. However oxidation of 73 by the

Swem method, as described in the literature,69 repeatedly gave only 10-40% mass recovery and 20-40% conversion to aldehyde 74.

The reaction conditions were adjusted by changing the DMSO:(COC1)2:ROH ratio from 6:3:1 to 4:2:1 and increasing the reaction temperature from -78°C to -60°C to give, in one case, a dramatically increased (80%) yield of aldehyde 74 after column Chapter 2 Results and Discussion 28 chromatography. Unfortunately, repetition of the reaction indicated that the yield of the

Swem oxidation of 73 remained variable, from 30-80%. The identity of the product as the aldehyde 74 was supported by 1H n.m.r. spectroscopic analysis even though the spectrum was made complex by the presence of diastereomers. In particular, two doublets were observed at 6 9.50 (J 1.5 Hz) and 9.64 (J 2.5 Hz) due to the aldehyde protons of the two diastereoisomers.

Pyridinium chlorochromate (PCC) has been used successfully to oxidize sensitive primary alcohols to aldehydes,70 but its use in this study as a possible replacement for the Swem oxidant for alcohol 73 was abandoned after it gave only 10% yield of the desired aldehyde 74.

Diisobutylaluminium hydride (DIBAL) has been used as a reducing agent to convert esters directly into the corresponding aldehydes at low temperature (-60°C). 71 •72

This presented a different approach to aldehyde 74, and one worthy of evaluation. Ester

72 was able to be reduced on 100 mg scale to aldehyde 74 with DIBAL in toluene solvent in 50% yield after purification (Scheme 16, Table I, Entry 1). However, when the scale was increased twenty-fold, the aldehyde 74 was obtained in only 20% yield.

There appeared to be a problem during the workup due to the low boiling point of the aldehyde 74 (53-S4°C/l 7 Torr)68 and comparatively high boiling point of the toluene solvent.

Thus, the reaction was repeated in dry CH2Ch at -78°C with DIBAL added as a toluene solution. Workup was also changed by quenching the mixture with potassium fluoride in a little water (Table I, Entry 2). The yield of aldehyde 74 was then increased from 20% in toluene to a more consistent 40-50% in CH2CJi. However, the use of Chapter 2 Results and Discussion 29

DIBAL in hexane (Entry 3) rather than in toluene solution, and a change in the workup to a simple water quench almost doubled the yield (89%) and provided good access to the aldehyde 74 on moderate scale; the reaction also became reproducible.

DIBAL Me,,...... _CHO

72 74

Conditions: see Table 1

Scheme 16

Table 1. Reduction of ester 72 to aldehyde 74 by DIBAL

Entry DIBAL Solvent Workup Yield%

1 toluene toluene aq. NaHS2O3, 20-50 NaOH 2 toluene CH2Cli sat. KF 40-50

3 hexane CH2Cli H2O 89

With the aldehyde 74 in hand, it was next subjected to Wittig olefination.

According to the literature67 stereoselective Wittig olefination of aldehyde 74 with

(ethoxycarbonylmethylene)triphenylphosphorane 75 in methanol afforded an 82:18 mixture of the (Z)- and (E)- pentenoates 76a and 76b, respectively, after column chromatography. Indeed, the product comprised four diastereoisomers, which were evident in the 1H n.m.r. spectrum through the appearance of two sets of doublets of doublets at 6 5.69, 5. 75 (J 11.8, 1.6 Hz) and 6.11, 6.22 (J 11.8, 8.2 Hz), due to the a and f3-olefinic protons in the (Z)-fonn 76a, respectively, and another two at 6 5.91, 5.96

(J 14.3, 1.0 Hz) and 6.83, 6.92 (J 15.9, 6.2 Hz), due to the a- and f3-olefinic protons in Chapter 2 Results and Discussion 30

the (E)-form 76b, respectively. Kang67 has reported the preparation of compound 76, but did not comment on the four stereoisomers. In the current experiments, the (Z)- and

(E)-isomers were present, which could not be noticeably improved upon chromatography. The mixture was therefore used in the next step without separation.

Treatment of the (Z)- and (£)- mixture 76 with a catalytic amount of 30% aqueous sulfuric acid in methanol presumably liberated the free alcohol which spontaneously underwent ring closure to form (+)-5(S)-methyl-2(5H)-furanone 30. The substance was isolated after chromatography as a pale yellow oil, [af; = +8.6°, c = 3.5,

CH3Cl, (lit.67 [ a]o = +8.3°, c = 3.3, CH3Cl) in 71 % yield.

This method therefore provided the Michael acceptor, lactone 30, in four steps rather than the five steps reported in the literature, 67 in an overall yield of 51 % compared to the reported in 4 7% yield.

2.4 Preparation of Michael Donors

The preparation of a suitable Michael donor for reaction with lactone 30 was the most challenging and frustrating part of this research. Inspection of Scheme 12 indicated that the simplest synthetic equivalent of aromatic 66 would be an arylacetaldehyde 77

(66, X = H, Y = O, Z = Br). There are many types of Michael donors reported in the literature, 73 but there is no precedent for aldehyde enolate addition to unsaturated lactones or esters.74 It was likely that self-condensation of the aldehyde in an aldol process would interfere. Hence a structural equivalent of the aldehyde was sought.

Carbanions derived from ketone and aldehyde hydrazones are highly nucleophilic and have been used as nucleophiles in conjugate addition.75'76 Moreover, they are Chapter 2 Results and Discussion 31

unlikely to yield products of self-condensation because the hydrazone group is not highly electrophilic. A N,N-dimethylhydrazone was actually used as a key reactant in the synthesis of simplified analogues of himbacine.43 In a related process, enamines are often sufficiently reactive to add to many conjugate acceptors by simply mixing and

(sometimes) heating the two together in solution.77 This reaction forms the basis of many cyclocondensation processes s.uch as the Robinson annulation.

Since hydrazones and enamines are most commonly derived from the corresponding carbonyl precursors, several attempts were made to prepare aldehyde 77.

The various approaches to compound 77 will be discussed in the following section.

2.4.1 Attempted preparation of 2-(2-bromomethylphenyl)ethanol 79

Oxidation ofindene 78

It was initially planned to synthesise aldehyde 77 from indene 78 as outlined in

Scheme 17. This route was thought flexible in the types of groups that could be attached to the aromatic ring.

ii ~OH CHO ------~ OC ~OH 00 - CHO

78 80 82

I ••• I Ill I t ~o iv ~OH ~Br ~Br

77 79

Reagents: i. o3, MeOH or CH2Cl2; ii. NaBH4; iii. HBr; iv. Swem oxidation

Scheme 17 Chapter 2 Results and Discussion 32

0 zono lys1s · 78-80 prov1'd e d a d'irect synthesis. of compound 80. Hence, indene 78 was treated with a stream of ozone in methanol at -30°C for 6 hours. However, reductive workup with dimethyl sulfide gave little diadehyde 80 as evidenced by the 1H n.m.r. spectrum of the crude product. Instead, significant amounts of what appeared to be acetals and probably hemiacetals, 84a, 85a and 86a, were detected and must have arisen from participation by the methanol solvent.

Thus indene 78 was also treated with ozone in dry dichloromethane for up to 6 hours at -65°C, following the method of Garratt and Vollhardt,79 however, zinc-acetic acid workup gave only recovered indene 78.

Alternatively, dialdehyde 80 has been obtained by oxidation of trans-1,2- indanediol with sodium periodate, solid-phase periodate82 or with lead tetraacetate.83•79

Oxidation of indene 78 with potassium permanganate was reported84 to yield diol 83 in

56%. In practice, treatment of indene 78 with aqueous potassium permanganate in THF

(Scheme 18), gave a crude product that was oily. The oil was distilled to give from the distillate residue, in 7% yield, a white crystalline solid, m.p. 107-108°C, whose elemental analysis and mass spectroscopic data (m/z 150 (M+)) were supportive of the diol 83. The 1H n.m.r. spectrum contained an ABX pattern of signals at B 2.93 (dd, J

16.3, 3.6 Hz), 3.09 (dd, J 16.3, 5.8 Hz) and 4.44 (brs) which were consistent with protons at C3 and C2, respectively, in diol 83. There was also a doublet of doublets at B

4.95 (dd, J 5.1, 5.0 Hz) that appeared to have arisen from the proton at Cl but required that spin coupling to the benzylic hydroxyl proton had occurred. In addition there were evident, two broad signals at B 2. 75 (J ea 4 Hz) and 2.87 (J 6 Hz) that could only be assigned to the two hydroxyl signals. The signals at B 2. 75 and 2.87 did disappear and Chapter 2 Results and Discussion 33

the broad signals at 6 4.44 and at 4.95 sharpened to doublet of doublet and doublet

signals, respectively, upon deuterium exchange. Therefore, the structure of diol 83 was

confirmed. The cis geometry of the compound 83 was established by NOESY

correlation experiments. The signals at 6 2.75 (OH) and 2.87 (OH) showed a strong

nuclear Overhauser enhancement (nOe) correlation to each other, although there was no

nOe correlation between the signals at 6 4.44 (H2) and 4.95 (HI).

iii

OR1 O)···110H ii co ~R'2 ''oH OR2 78 83 91 R2 84a H Me 85a Me H 86a Me Me 84b H Et 85b Et H 86b Et Et

Reagents: KMn04, THF/H20; ii. KI04, MeOH; iii. 03, EtOH

Scheme 18

It was felt that the low yield of diol 83 might have resulted because there was low conversion from indene 78. The oxidation was repeated under more vigorous conditions by extending the reaction time to 6 h at r.t. and by heating the mixture to 40-

500C for 3 hours and 6 hours. The diol 83 was the only product recovered from all these experiments after distillation. The best conditions were determined to be reaction at 40-

500C for 3 hours, but in each case the yield was low (10-12%). Chapter 2 Results and Discussion 34

Although only very small quantities of material were at hand, the diol 83 was next treated with sodium periodate in 95% ethanol containing a trace of SM H2S04•

Workup and analysis of the 1H n.m.r. spectrum of the crude product revealed only two minor signals at o 9.9 and 10.0 due to the presence of aldehydes. The major product was not aldehydic but contained ethoxy signals that, from the non-equivalence of their methylene signals, were diastereotopic. The crude material decomposed during attempted purification on silica gel but the major product was tentatively assigned as a mixture of -/t8oh"ate5 and acetals 84b, 85b and 86b in which the solvent had participated in reaction following oxidation. Closer examination of the literature79 did reveal mention of the formation of such product when diol 83 was oxidized in ethanol.

Indeed, the literature85 went on to describe oxidation of cis-indanediol 83 using sodium periodate in moist tetrahydrofuran, whereupon the dialdehyde 80 was afforded in 99.9% yield. Thus, the diol 83 was treated with sodium periodate in aqueous tetrahydrofuran ( 1: 1) in the presence of 2M H2S04• Workup gave excellent mass recovery of an oil whose 1H n.m.r. spectrum contained a singlet at o 10.03 and a triplet at 9.80 (J 1.6 Hz). These signals each integrated for one proton and indicated that the dialdehyde 80 was the major product. Unfortunately, the sample polymerized within 3 days when stored in the freezer, a result that was noted in the literature79 for samples that were impure.

The preparation of bromoaldehyde 77 from indene 78 was clearly problematical, and a more radically different approach was sought. Chapter 2 Results and Discussion 35

via Claisen rearrangement ofally/ ether 101

It was recognised that additional functionality would eventually be needed in the aromatic Michael donor to introduce radiolabels into the target molecules. Moreover, it was considered that an hydroxyl group might be suitable as a vehicle for this functionality. Hence the synthesis described in Scheme 19 was investigated. This pathway should enable the sensitive acetaldehyde group to be introduced late in the synthesis through cleavage of an allyl appendage. Advantage was taken of the additional oxygen substituent to introduce the C-allyl group through Claisen rearrangement of an allyl ether.

ii // i • -----~ (00 HO 0 CH3 HO0 CH2Br -' CH2Br

88 90 87

I ••• j Ill I t

V iv ~o .., ___ ~---- PO CH Br HO0::: CH2Br PO ~I " 2

93 92 91

Reagents: i. NBS, CCl4; ii. allyl bromide 89, acetone; iii. heat; iv. protection; v. 03

Scheme 19

Thus, m-cresol 88 was treated at reflux with N-bromosuccinimide (NBS) in carbon tetrachloride, in the presence of benzoyl peroxide, as the mixture was irradiated with white light. The intention was to perform free radical bromination of the aryl methyl group. In practice, halogenation occurred on the 2, 4 and 6 positions of the Chapter 2 Results and Discussion 36

benzene ring (Scheme 20); none of the desired 3-bromomethylphenol 90 was detected.

The reaction gave a colourless oil in quantitative yield. Analysis of the 1H n.m.r. spectrum revealed the oil to comprise three components through the appearance of three methyl singlets at 6 2.29, 2.33 and 2.41 in a 1:1.5:1 integral ratio. It was noted that if bromination had occurred on the methyl group, the chemical shift of those signals would have been higher. The n.m.r. signals for the aromatic protons also appeared in distinctive patterns that indicated the three types of substitution in structures 94, 95 and 96.

Thus, one set of aromatic proton signals at 6 6.81 (d, J 7.7 Hz), 6.87 (d, J 6.7

Hz) and 7.11 (dd, J 7.7, 7.7 Hz) were indicative of 1,2,3-trisubstitution, as found in compound 94. Another two sets of aromatic proton signals at 6 6.55 ( dd, J 8. 7, 3.1 Hz),

6.73 (d,J3.1 Hz), 7.35 (d,J8.7Hz)and6 6.62 (dd,J8.2, 2.1 Hz), 6.85 (d,J2.l Hz),

7.32 (d, J 8.2 Hz) indicated the presence of two 1,2,4-trisubstituted phenols. The two narrow doublets at 6 6. 73 and 6.85 could only have arisen from protons with only meta coupling and the two wide doublets at 6 7.35 and 7.32 were too close in chemical shifts to allow a distinction to be made between structure 95 and 96. It was concluded that the smaller chemical shift of the doublet of doublet signals (o 6.55) was due to H6 in compound 95, since this proton should be the more shielded due to the neigbouring hydroxyl group. Chapter 2 Results and Discussion 37

ri + OBr + Br"("'-i HO0 CH3 HO~CH3 HO CH3 HO~CH3 Br 88 94 95 96 Iii

~ + AcOV CH3 AcO~CHBr2

97 98 99

Reagents: i. NBS, CCl4; ii. (CH3CO)2O, pyridine

Scheme 20

It appeared that the aromatic ring of m-cresol was too highly activated towards

bromination by the OH substituent. Therefore, m-cresol was first acetylated and the

bromination repeated (Scheme 20). The acetate 97, obtained in 86% yield, was then w~ treated with NBS in carbon tetrachloride to afford after distillation the desired

compound, 3-(bromomethyl)phenyl acetate 98 in 74% yield, and the higher boiling,

doubly brominated product, 3-( dibromomethyl)phenyl acetate 99 in 10% yield.

Compound 98 gave the expected elemental analysis results and isotopic 1: 1 molecular

ions at mlz 230 and 228 in its mass spectrum. The bromomethyl group was evident in

the 1H n.m.r. spectrum through a shift in the methyl singlet from 6 2.38 in m-cresol to

4.47, in keeping with the more electronegative environment. The dibrominated

gfl, · . I I . . 1 2 1 1. compound 99 was observed to have three 1sotop1c mo ecu ar tons m a : : ratio. at m,z

310, 308 and 306 in its mass spectrum due to the presence of two bromine atoms in the molecule, and fragmentation ions at mlz 267, 265 (base peak) and 263 (ratio 1:2: 1)

corresponding to loss of an acetyl radical from the three parent isotopic ions. There was Chapter 2 Results and Discussion 38

also evident in the 1H n.m.r. spectrum a methine signal resonating at 6 6.62 due to the

dibromomethyl group. Integration of the respective bromomethyl and dibromomethyl

signals permitted the molar ratios of the two compounds in mixtures to be calculated.

Attempts to prevent over reaction, by reduction of the reaction time, slower,

portionwise addition of the NBS, or reduction of the molar equivalents of NBS (from

1.15 to 1.00) failed to completely stop the occurrence of the more highly brominated

compound. The optimum conditions for monobromination were deduced to be combination of the acetate 97, NBS, and benzoyl peroxide in a mole ratios of I: 1.1 :0.2 at reflux for 3 hours.

In continuation of the synthesis, compound 98 was treated with 10% potassium hydroxide in methanol to form the phenolic salt. Evaporation to dryness and treatment of the residue with allyl bromide in acetone was expected to give 0-allylation (Scheme

21, Route A). A new product was isolated, but it was not the desired allyl ether 87.

There was no evidence in the 1H n.m.r. spectrum of allyl resonances. Instead, the singlet at 6 4.47 from the bromomethylene protons had been replaced by two new singlets at

6 3.40 (3H) and 4.44 (2H). There also appeared an exchangeable signal at 6 6.27 that was consistent with a phenolic proton. These results indicated that the ester group in 98

had been cleaved, but that the bromine on the methylene group had been replaced by a

methoxy group, probably through participation of the methanol solvent in nucleophilic

~f,(1, substitution to give the phenol 100. The identity of compound 100 was confirmed by

analysis of its infrared, n.m.r. and mass spectra. It gave a broad absorption at 3320 cm·1

in the infrared spectrum due to the hydroxyl group in the molecule, and a very strong

molecular ion at mlz 138, with no evidence of bromine isotopes in its mass spectrum. Chapter 2 Results and Discussion 39

The aromatic splitting pattern in the 1H n.m.r. spectrum was also consistent with

retention of the meta-disubstitution of the aromatic ring.

Route A

Di ii AcO A? CH2Br [ +K·O°'CH2R ] HO 0 A? CH2OMe 98 R=OCH3 100 or Br

Route B

iii ii HO 0 CH2OMe (00 A? CH2OMe 100 101

Reagents: i. 10% KOH, MeOH; ii. allyl bromide 89, K2CO3, acetone; iii. 2M HCI, MeOH, reflux

Scheme 21

In support of the hypothesised mechanism of formation, compound 98 also gave the methoxymethylphenol 100 when subjected to treatment with 2M hydrochloric acid in methanol (Scheme 20, Route B). The products from the two pathways were identical in their chromatographic and spectroscopic characteristics.

Advantage was taken of this result. Compound 100 was treated with allyl bromide in the presence ofK2C03 in dry acetone to give the methoxymethyl substituted allyl ether 101. The structure of allyl ether 101 was supported by the infrared spectrum which showed a strong stretching vibration at 1585 cm·1, corresponding to the alkene group and no hydroxyl stretching absorption. The 1H n.m.r. spectrum gave two doublets at 5.29 (J 10.8 Hz) and at 5.42 (J 17.4 Hz), corresponding to the two terminal vinyl Chapter 2 Results and Discussion 40 protons, and a multiplet at O 6.06, corresponding the remaining olefinic proton. The methylene protons from the allyl group resonated at o 4.55, which was consistent with their position next to oxygen.

Allyl ether 101 appeared relatively stable and was subjected to Claisen rearrangement conditions.

When the allyl ether 101 was heated for 6 hours in N,N-dimethyaniline at 185-

1900C, relatively standard conditions, only allyl ether was recovered (Scheme 22, Table

2). The ether 101 was therefore sealed in an ampoule without solvent and heated to 190-

2000C in a Woods metal bath. Three rearrangement products, presumably 102, 103 and

104, were observed, as evidenced by the appearance of three sets of 1H n.m.r. signals at o 4.40, 4.42 and 4.45 in the ratio 50: 10:40, respectively. These signals were later assigned to the methylene protons on the methoxymethyl group. There was also er observed three broad singlet signals at 4.81, 5.49 and 5.62 corresponding to hydroxyl protons. Column chromatography on silica gel afforded two isomers in relatively pure form. The first isolated substance gave in its 1H n.m.r. spectrum three aromatic proton signals, at o6.77 (d, 18.2 Hz), 6.93 (d, 17.2 Hz) and 7.10 (dd, 17.7, 7.7 Hz), that were indicative of 1,2,3-trisubstitution. Therefore, the substance was identified as compound

102. The second substance also gave three aromatic proton signals at O 6.81 (s), 6.84

(dd, 17.7, 1.6 Hz) and 7.08 (d, 17.7 Hz). The singlet could only have arisen from an isolated proton in the aromatic ring, and indicated the presence of 103 or 104.

Meanwhile, the doublet at o6.84 could have arisen either from the proton at position 4 in compound 103 or at position 6 in compound 104. Based on its low chemical shift, this signal was assigned to the proton at position 6 in 104 since only this proton should Chapter 2 Results and Discussion 41 be shielded by a neigbouring hydroxyl group. Thus, the second substance was assigned to the desired para-rearrangement product 104. The remaining product was therefore assigned to isomer 103.

heat ~ ~ (QI -HO1c~ # OMe O # OMe .., + HO~OMe + HO~OMe

101 102 103 104

Scheme 22

Table 2. Claisen rearrangement of allyl ether 101

Solvent Temperature Time, h Starting Product ratio oc material% 102:103:104 dimethyl 185-190 0/N 98 trace aniline none 220 0/N trace

none 195-200 3 30 25:25:50

none 195-200 4 0 50:12:38

none 195-200 6 0 50:10:40

Repetition of the reaction for 3 hours, 4 hours and 6 hours, respectively, initially favoured the ortho isomer but thereafter afforded minor variations in the proportion of isomers (Table 2). When the reaction time was prolonged to 6 hours, there was evidence that polymeric product was formed. Apparently, there was no increase in the proportion of para-rearrangement product 104 with increasing vigour of the reaction.

Thus, the Claisen rearrangement approach to aldehyde 93 appeared unsuitable for synthetic purposes. Chapter 2 Results and Discussion 42

Before this route was entirely abandoned, a more direct synthesis of a C

allylated precursor to aldehyde was examined. 3-Chloromethylanisole 105 was

commercially available and was being used in the laboratory for another project. It was therefore treated with allyl chloride in chlorobenzene in the presence of a stoichiometric amount of aluminium chloride (Scheme 23). These Friedel-Crafts conditions were expected to yield predominantly the product of allylation para to the more strongly activating methoxy group, namely compound 106. Unfortunately, no reaction took place after 1 hour at 40°C and aqueous workup gave only unreacted 3-chloromethylanisole

105.

I') ~ MeO~CI MeO~CI

105 106

Reagents: i. allyl chloride, AICl3, PhCI, 40°C

Scheme 23

Another, entirely different pathway to aldehyde 77 was sought.

The ready availability of a,a'-dibromo-o-xylene 107 prompted a brief approach to bromoaldehyde 77 involving selective functional group manipulation. It has been noted on many occasions that substitution of the halogen group of 107 can be a.ciw.v.ed wi:fi .f,ase...... Sttestittttcd. For example, reaction of a,a-d1bromo-o-xylene 107 with the potassmm salt of pyrrole under different reaction conditions led to the symmetrical bis-pyrrolyl derivative through substitution. 86 Another group has reported87 that reaction of 4- chloro-a,a' -dibromo-o-xylene with 2.5 molar equivalents of sodium cyanide (NaCN) gave 4-chloro-1,2-bis(cyanomethyl)benzene, the result of two-fold substitution. Chapter 2 Results and Discussion 43

• JI . ~ w'"t-11.-. ~u4 \. 5v-2-1v1s However, selective mono-substitution of the halogens of compound 107 lhas R'*- been -

4 reported in the literatur,~1:, ~ k- r.M,~·CC-e-('S"-hl!ai~l ~ r-~ VoJe-! ~~ i.-,. ~ ~ ~~e,+1 o./- wi,,.;-h,-kt.-hC.:...... ~ tZ~ ~ ~{,.~~. Despite the lacking of literature precedents, dibromide 107 was treated with one

equivalent ofNaCN in a mixture of benzene and methanol (2:1) at room temperature for

one day. Aqueous workup gave a 61:39 mixture of dibromide 107 and dinitrile 109 in

48% yield (Scheme 24). An authentic sample of the dinitrile 109 was prepared by treating the a,a'-dibromo-o-xylene 107 with four equivalents ofNaCN and the product

109 was isolated in 97% yield. Thus, it appeared that selective displacement of benzylic halides would not provide a pathway to aldehyde 77 and effort was directed towards yet another route that showed more promise.

~Br CN CN ~Br OCBr OCCN

107 108 109

2.4.2 Preparation of 2-(bromomethylphenyl)acetaldehyde 77

A successful synthesis of bromoaldehyde 77 was finally achieved from isochroman 110.

It was reported88 that isochroman 110 with gaseous hydrogen bromide gave 2-(2- bromomethylphenyl)ethanol 79, a potential precursor to aldehyde 77, in 63.5% yield. In the absence of a cylinder of hydrogen bromide, the isochroman was treated with more readily accessible gaseous hydrogen chloride, but there was no reaction. It was possible that the combination of weaker acid and less nucleophilic halide anion prevented reaction. Isochroman 11 0 was then treated with excess 48% aqueous hydrobromic acid for different reaction times and at different temperatures (Table 3). Most of the Chapter 2 Results and Discussion 44

conditions gave only the starting material 110. However, a warmed reaction mixture

(Table 3, entry 3) gave a 5% yield of a new substance, later identified as the dibromide

111. This result gave some encouragement but revealed a lack of selectivity in the substitution reaction. The stronger acid, hydroiodic acid, which is a classical reagent for ether cleavage, was also used to treat isochroman 110. Reaction at ambient temperature for 46 hours gave what was believed to be iodoalcohol 112 in 8% yield. Compound 112 was evident in the 13C n.m.r. spectrum through the appearance of a strong upfield carbon signal at o 4.3, but the compound was unstable and decomposed during attempted purification.

Finally, anhydrous hydrogen bromide was generated by treatment of tetralin with bromine.89 The isochroman 110, in the absence of solvent, was saturated with hydrogen bromide gas and the mixture kept at ambient temperature for 17 hours.

Aqueous workup and chromatography on silica gel afforded the desired bromoalcohol 79 in 68% yield, unreacted isochroman 110 in 10% yield, and the less polar dibromide 111 in 4% yield. The presence of bromine in the bromoalcohol 79 was most evident from the mass spectrum where there was observed isotopic molecular ions at m/z 216 and 214.

The base peak at mlz I 04, corresponded to loss of bromine and a CH20H radical from the parent molecule. When the reaction time was prolonged to 24 hours, there was no increase in the proportion of product 79. Indeed, there was evidence that the dibromide

111 was increased. Chapter 2 Results and Discussion 45

~OH ~Br CX) or ii ~x + ~Br

110 79 X = Br 111 112 X = I

Reagents: i. HBr (g), r.t., 17 h; ii. HI (55%), r.t. 46 h.

Scheme 24

Table 3. Reactions of isochroman 110

Entry Acid Temperature Time, h Yield% oc 110 other 1 HCl (g) r.t. 17 75

2 HBr (48%) r.t. 6-29 80

3 HBr (48%) 50 24 55 5 (111)

4 HBr (48%) 50 7 65 trace (111)

5 HI (55%) r.t. 46 8 (112)

6 HBr(g) r.t. 17 10 68 (79) 4 (111)

Formation of the dibromide 111 was unwanted, but generally the synthesis was efficient and above all, reproducible.

With the bromoalcohol 79 now accessible in reasonable amounts, it was treated with 2.5 mole equivalents of pyridinium chlorochromate in dichloromethane (CH2Ch) to afford aldehyde 77 in 45% yield (Scheme 25). The infrared spectrum of the product gave a carbonyl stretch at 1720 cm·1• The expected molecular ion was absent in the electron impact and electrospray mass spectra, but a base peak at mlz 133 corresponding to loss of a bromine radical from the molecule was observed. In the 1H n.m.r. spectrum, compound 77 gave a narrow triplet at 6 9.80 (J 2.1 Hz), expected from the aldehydic Chapter 2 Results and Discussion 46

proton, and a matching two-proton doublet signal at 3.88 (J 2.1 Hz) corresponding to the methylene protons next to the aldehyde group; in the 13C n.m.r. spectrum, there was a signal at o 198.3 which supported the presence of an aldehyde carbonyl group. There also remained a two-proton singlet corresponding to the bromomethylene protons in a similar position to that found in the bromoalcohol 79. Although efforts were made to fully characterise this substance by elemental analysis, it appeared unstable during column chromatographic purification.

~OH ~o ii 113 ~Br ~Br

79 77

114

Scheme 25

Interestingly, pyridinium dichromate (PDC) and Swem oxidations of alcohol 79 were not successful. Therefore, although the yield of aldehyde 77 from pyridinium chlorochromate reaction was not high, it was thought acceptable at this stage.

Attention was next turned to the conversion of aldehyde 77 into its N,N dimethylhydrazone 113 in readiness for the planned Michael addition reaction with lactone 30. Surprisingly, when aldehyde 77 was treated with 1.0 mole equivalent of N,N dimethylhydrazine in dichloromethane, the desired hydrazone 113 was formed in only trace amount and could not be isolated (Scheme 25). A major new product was obtained Chapter 2 Results and Discussion 47

but the oily substance was tentatively assigned as amino hydrazone 114 (47% yield).

The new compound gave two six-proton singlets at 6 2.71 and 3.46 in the 1H n.m.r.

spectrum suggestive of the presence of two dimethylamino groups. There also appeared ,r a two-proton doublet at"-3.77 (J 5.6 Hz) and a one-proton triplet at 6.56 (J 5.6 Hz) that

were supportive of the acetaldehyde hydrazone portion of the molecule. A two-proton

singlet at 6. 74 replaced the higher field signal for the bromomethyl group and it strongly

indicated that the bromine had also been replaced by a dimethylamino group.

This result was surprising but in keeping with the ready substitution of benzylic

bromides that led to formation of methoxymethylphenol 100 (see Scheme 21).

The wayward reaction of aldehyde 77 with N,N-dimethylhydrazine and therefore the unsuitability of 77 in the long term synthesis of intermediate 65, and

ultimately himbacine analogue 64, was extremely disappointing. Fortunately, the

flexibility of the synthetic scheme enabled a modified co1.1rse to be taken.

2.4.3 Preparation of 2-(2-methoxymethylphenyl)acetaldehyde N,N

dimethylhydrazone 117

Advantage was taken of the high reactivity of the benzylic halide to transform

bromoalcohol 79, through treatment with methanol at reflux, into methoxyalcohol 115

(Scheme 26). The reaction proceeded in high yield and with few side products. About

I 0% of the isochroman 110 was regenerated during the reaction but it could be recovered

and reused. Compound 115 gave a small molecular ion at mlz 166 in the mass spectrum.

Importantly, there appeared a three-proton singlet at 6 3.41 in the 1H n.m.r. spectrum Chapter 2 Results and Discussion 48

due to the methoxy group, and the benzylic methylene signal from bromoalcohol 79 at B

4.58 was shifted to B 4.47 in compound 115.

cx:;;°H cc:OH + r Me 00 79 115 110

ii !

iii cc:oMe

116 117

Reagents: i. MeOH, reflux, 3 h; ii. PCC, CH 2Cl2, r.t.; iii. H2NNMe2, EtOH, reflux

Scheme 26

Since the bromine in compound 79 was eventually replaced by a methoxy group, a potentially. shorter route to alcohol 115 was briefly investigated. Isochroman was treated in methanol with a little perchloric acid in an attempt to form compound 115 directly (Scheme 27). However, when the mixture was refluxed for 48 hours, only isochroman 110 was recovered. This outcome was reminiscent of the results from earlier attempts to convert isochroman 110 to bromoalcohol 79. It therefore appeared that the ring closed form of isochroman was relatively stable and the presence of hydroxylic solvents discouraged ring opening. The isochroman impurity from the reaction in

Scheme 26 therefore might have arisen from reversal of the ring opening process.

~OH -----~ 00 ~OMe 110 MeOH 115

Reagents: i. HCI04, MeOH, reflux

Scheme 27 Chapter 2 Results and Discussion 49

While the methoxy group was a poorer leaving group than bromide, it was

thought a useful stabilising influence in order to test later synthetic steps and

sufficiently strong a leaving group to participate in the final cyclisation process. Thus,

the synthesis was continued.

Alcohol 115 was oxidized with pyridinium chlorochromate (Scheme 26), the

oxidant found suitable for conversion of bromoalcohol 79 into the corresponding aldehyde 77, to give aldehyde 116 in 85% yield. On this occasion when aldehyde 116 was subsequently treated with 3.0 mole equivalents of .N,N-dimethylhydrazine in ethanol, the reaction afforded the desired hydrazone 117 as a yellow oil in 87% yield.

This confirmed that the undesired reaction of bromoaldehyde 77 with .N,N dimethy lhydrazine was due to the reactivity of the benzylic halide. Compound 117 gave the expected infrared absorption at 1595 cm·1, due to C=N stretching and its elemental analysis results were consistent with the molecular formula C12H18N20. The hydrazone

feature was most evident in the 1H n.m.r. spectrum where there was a six-proton singlet

at 6 2.75 corresponding to the .N,N-dimethyl group and a one proton triplet at 6 6.65

resulting from the aldimine proton.

2.4.4 Preparation of 2-(2-phenoxymethylphenyl)acetaldehyde 120 and its

hydrazone 121

As mentioned earlier (section 2.4.3), the methoxy group was not an ideal leaving

group. Therefore, efforts were made to synthesise analogues of compounds 115, 116

and 117, in which the methoxy substituents were replaced by better leaving groups,

such as trifluoroethoxy and phenoxy groups. Chapter 2 Results and Discussion 50

Bromoalcohol 79 was first refluxed in trifluoroethanol for 3 hours, but instead of the desired nucleophilic substitution occurring (Scheme 28), the reaction gave an 85: 15 mixture of bromoalcohol 79 and isochroman 110. This result indicated that the intramolecular reaction proceeded faster than the desired intermolecular substitution ' even though the nucleophile was used as solvent.

~OH ~OH i or ii ~Br ~A

Reagents: i. CF3CH2OH, reflux; ii. PhOH, K2CO3 , acetone

Scheme 28

A range of mild bases might have been used to permit deprotonation of the trifluoroethanol and thereby encourage the intermolecular process. Instead, phenol and bromoalcohol 79 were reacted together in acetone in the presence of anhydrous potassium carbonate. The reaction afforded the phenoxymethyl derivative 119 as a colourless oil in 62% yield. The substance gave the expected molecular ion at m/z 228 in the mass spectrum and a base peak at mlz 94 corresponding to phenol, a likely fragment.

In the 13C n.m.r. spectrum, there were also four additional carbon signals at 114.8, 121.2,

129.6 and 158.6, corresponding to the newly introduced phenoxy group.

~OH ~o ii ~OPh vl--,OPh

119 120 121

Reagents: i. PCC, CH2Cl2, r.t.; ii. H2NNMe2, EtOH, reflux.

Scheme 29 Chapter 2 Results and Discussion 51

Oxidation of the new phenoxyalcohol 119 with pyridinium chlorochromate gave the desired aldehyde 120, as a pale yellow oil in 38% yield (Scheme 29). Its structure was shown by the appearance in the 1H and 13C n.m.r. spectra of signals at 6 9.77 (t. J

2.1 Hz) and at 199.3, corresponding to the aldehyde proton and carbon, respectively, and at 6 3.80 (d, J 2.1 Hz) and 47.8, corresponding to the neighbouring methylene group. In addition, the two methylene resonances at 6 2.98 and 3.91 in the spectrum of compound 119 had disappeared. The aldehyde 120 seemed less stable than the methoxy aldehyde 116 during chromatography and could not be made analytically pure.

However, treatment with N,N-dimethylhydrazine afforded the desired hydrazone 121 as a more stable entity in 63% yield after column chromatography. The mass spectrum showed a molecular ion at mlz 268 and a strong fragment ion at mlz 175 corresponding to loss of the phenoxy radical from the parent molecule. Elemental analysis confirmed the molecular formula C17H20N20. The 1H n.m.r. spectrum gave a new singlet at 6 2.73 and a triplet at 6.69 (J 5.6 Hz), that corresponded to the two new equivalent N-methyl groups and the aldimine proton, respectively. The assignment of the product as compound 121 therefore was assured.

~OH ~o ~OR- ~OR

116 R=Me 79 115 R=Me 110 119 R=Ph 120 R=Ph ! ~NNMe, R

117 R=Me 121 R=Ph

Scheme 30 Chapter 2 Results and Discussion 52

In summary, in this phase of the project methods had been developed for the synthesis of two Michael donors, each in four steps as outlined in Scheme 30.

2.5 Michael Reactions

The Michael reaction constitutes one of the most important preparative methods in organic chemistry and has a relatively long history.90•92 It involves the addition of a nucleophilic carbon species (Michael donor) to an electrophilic multiple bond (Michael acceptor) and is a particular class of the more general conjugate addition reaction.93•95

The reaction is applicable to a wide variety of anionic, e.g. enolate, nitronate and phosphonate, and neutral, e.g. enamine, donors. For example, Croft and his co-workers reported96 that enamine derivative 122 underwent reaction with methyl vinyl ketone, and subsequent hydrolysis, yielded a dione 123 in good yield (Scheme 31 ).

0 ~ i, ii

122 123

Reagents: i. dioxane, 16 h; ii. NaOAc, HOAc, H20, reflux

Scheme 31

Meanwhile, the electrophilic partners are typically a,~-unsaturated ketones, esters, or nitriles, nitroalkenes and unsaturated phosphonates, but many electron withdrawing substituents can also activate carbon-carbon double bonds to nucleophilic attack. Alkynes can also participate as Michael acceptors. Chapter 2 Results and Discussion 53

An example of the Michael reaction that is particularly relevant to this project is outlined in Scheme 32.97 Formation of the vinylogous enolate anion from 124 followed by Michael addition and subsequent ring closure afforded a mixture of only two diastereomers 125a and 125b in the ratio of 83: 17. A high degree of diastereoselectivity was observed in the reaction, but importantly, both isomeric products had cis geometry about the ring fusion to the lactone. This demonstrated the control of stereochemistry offered during cyclisation once the stereocentre generated through conjugate addition was established.

Me9 H . - r"r'OMe i, ii + 0 ~Ar . 0 Ar OH 0

124 125a 125b

Reagents: i. LDA, THF, -78°C; ii. furan-2(51-1)-one, -78°C

Scheme 32

There was ample precedents for this type of Michael reaction in the literature on lignan synthesis. 98•101 One of the more relevant reports was by the Medarde group, 100 who synthesized podophyllotoxin 126 by means of a ciotni'rIO Michael addition alkylation sequence on 5H-furan-2-one, followed by cyclisation and controlled epimerizations (Scheme 33). 100 Chapter 2 Results and Discussion 54

st\ s 0 st""\

128 129

0 (°~o 0~ (') 0

MeoyoMe OMe

126 130

Reagents: i. Buli, THF, -78°C; ii. 5H-furan-2-one; iii. 3,4,5-trimethoxybenzaldehyde; iv. SnCl4, C~Cl2; v. HgO, BF3.OEt2 , THF; vi. AcOH, EtOH

Scheme 33

In the present study, the strategy for assembling the tricyclic portion of analogue

64 required stereoselective addition of a nucleophilic group through a Michael reaction to (5S)-methylbutenolide 30. It was envisaged that the Michael addition would take place predominantly from the sterically less hindered upper face, away from the 5- methyl substituent on the lactone 30.

2.5.1 Reactions of enamine 131 with lactone 30

Enamines are very specific and useful enol equivalents for aldehydes. 102 They react with electrophiles and add to Michael acceptors to give selective C-C bond Chapter 2 Results and Discussion 55

formation. 103 Hydrolysis of the resulting alkylated enamine can also regenerate the carbonyl group thereby making this an attractive route to intermediates for the synthesis of tricycle 65.

Accordingly, (EJ-N-~-styrylpyrrolidine 131 was prepared from phenylacetaldehyde 132 and pyrrolidine. The traditional method of synthesis, 104 in which the two reactants were refluxed together in toluene in the presence of an acid catalyst with removal of water, gave the enamine 131 in a satisfactory yield (85%).

However, an alternative method, involving treatment of phenylacetaldehyde 132 and pyrrolidine in ethanol in the presence of 4A molecular sieves, afforded the enamine 131 in almost quantitative yield (98%) with noticeably higher purity after Kugelrohr distillation.

ii do no reaction VoI I - ' 132 131

I iii

~0~ or

OTMS OTMS

133a 133b

Reagents: i. pyrrolidine,.4A molecular sieves, EtOH, reflux; ii. lactone 30, CH2Cl2; iii. lactone 30, TMSCI, CH2Cl2

Scheme 34 Chapter 2 Results and Discussion 56

The enamine 131 was isolated as a pale yellow oil that had identical 1H and 13C

n.m.r. spectroscopic properties to those in the literature. 104 It gave two mutually

coupled one-proton doublet signals (o 5.12 and 7.06, J 13.9 Hz) in the 1H n.m.r.

spectrum. The large coupling constant indicated that the alkenyl group was in the £

configuration. The substance also gave a molecular ion and a base peak at mlz 173 in the mass spectrum, which was consistent with the molecular formula C 12H 15N.

Simple combination of enamine 131 and lactone 30 in stoichiometeric quantities in CH2Ch within 6 hours at r.t. gave no reaction. When the enamine 131 was added to a solution of lactone 30 in CH2Cl2 at r.t., in the presence of trimethylsilyl chloride

(TMSCl), the product gave a new set ofn.m.r. signals that were eventually identified as coming from a product of reaction between enamine 131 and TMSCl alone (Scheme 34).

A mixture of enamine 131 and lactone 30 was also treated with trimethylsilyl trifluoromethanesulfonate (TMSOTf) in CDC13, but this gave a very complex mixture by 1H n.m.r. spectroscopy, and the reaction was eventually abandoned in favour of a carbanionic approach, as described in the next section.

2.5.2 Reactions of hydrazone derivatives

In order to test the Michael addition of hydrazone anions and gain experience in

this process, the reaction of N,N-dimethylhydrazone 134 with methyl crotonate was

examined. Hydrazone 134 was prepared as a model compound from N,N

dimethylhydrazine and phenylacetaldehyde 132 in absolute ethanol in 99% yield after

Kugelrohr distillation (Scheme 35). Chapter 2 Results and Discussion 57

Vo __ VNNMe, _;;,_;;_; -

132 134 135

Reagents: i. H2NNMe2, EtOH; ii. BuLi, Cul, THF, -78°C; iii. methyl crotonate

Scheme 35

It was then treated with n-BuLi to prepare the carbanion and with 0.2 mole equivalent of copper(!) iodide in order to encourage Michael addition. The reaction with methyl crotonate gave two diastereomeric products 135 in the ratio of 65:35 albeit in only combined 11 % yield (Scheme 35). The mixture was not separated but was characterised through its mass spectrum, which showed a molecular ion at mlz 262, and microanalytical results that were consistent with the molecular formula C15H22N20 2•

The presence of the two isomers and their ratio were evident from the appearance of two sets of doublet 1H n.m.r. signals at o 6.64 (J 7.4 Hz) and 6.68 (J 7.4 Hz), corresponding to the aldimine proton. In addition, there were duplicate signals throughout the 13C n.m.r. spectrum. The assignment of relative configuration in the two isomers was not possible with such a flexible, acyclic system.

Based on the experience of the above reaction and despite the low yield, hydrazone 134 was next treated with lactone 30 (Scheme 36) under the conditions used for reaction with methyl crotonate. The reaction on this occasion gave two diastereomers 136 in the ratio of 88:12 and in improved yield (18%). The presence of two one-proton doublet signals at o6.59 (J 5 .1 Hz) and 6.14 (J 3 .1 Hz) for the aldimine

(CH=N) proton indicated the presence of two diastereomers. In addition, the compound Chapter 2 Results and Discussion 58

gave in the mass spectrum a molecular ion at mlz 260, consistent with the expected product, and a base peak at mlz 161, due to the N,N-dimethylphenylhydrazone fragment with loss of the lactone. The microanalytical results were consistent with the molecular formula C1sH20N202. Thus the compound 136 was confirmed.

i, ii

0 134 R = H 136 R = H 117 R = CH20Me 139 R = CH20Me

Reagents: i. Buli, Cul, THF, -78°C; ii. butenolide 30

Scheme 36

The low yield of the desired products in the above two reactions was of concern, therefore another model system was studied. The more complex hydrazone 117 was made to react with methyl crotonate (Scheme 37) under the conditions used for hydrazone 134. Workup afforded a mixture that from 1H n.m.r. spectroscopic analysis, comprised recovered hydrazone 117 (9%), the desired addition product 137 (72%) as a

78:22 mixture of diastereomers (measured from integration of two aldimine proton (H5) doublets at o 6.51 and 6.57), and a third substance (19%). The mixture of two diastereomers was separated by column chromatography on silica gel to afford the major component, ester 137a, a single isomeric form as a pale yellow oil in 33% yield. The product 137a gave a molecular ion at m/z 306, a fragment ion at mlz 205 and a base peak at mlz 173, corresponding to the parent, loss of the methyl ester moiety (CsH902) and then loss of a methoxy group respectively. Elemental analysis results were consistent Chapter 2 Results and Discussion 59

with the molecular fonnula C11H26N203, and the infrared spectrum showed a carbonyl

absorption at 1725 cm· 1 and an imine stretch at 1600 cm· 1• The 1H and 13C n.m.r. spectra of compound 137 gave similar resonance patterns to those from compound 135 with the exception of the appearance of an AB quartet at o 4.42, 4.61 (J 11. 7 Hz) for the methylene protons of the methoxymethyl substituent and the presence of an additional methoxy singlet at 3.36 from the same group. As for hydrazonoester 135, the assignment of relative stereochemistry in the two isomers of 137 could not be made unambiguously.

~NNMe2 __i,_ii __

~OMe

117 137 138

Reagents: i. Buli, Cul, THF, -78°C; ii. methyl crotonate

Scheme 37

The third substance was obtained from the EtOAc-light petroleum (20:80) fraction as a pale yellow oil in 12% yield. The substance showed a molecular ion at m/z

332, of higher mass than the expected product, and a number of fragment ions at m/z

300, 201, 188, 173 and 131 in the electron impact mass spectrum. High resolution electrospray mass spectrometry gave an intense sodio-molecular ion at mlz 355.2386 which was consistent with a fonnula C20H32N20 2Na. The 1H n.m.r. spectrum was

similar to that of ester 137 but revealed the presence of a butyl chain (o 0.90, t, J 7.2

Hz, 3H), (1.29, m, 4H), (2.38, t, J 7.5 Hz, 2H), and loss of the methyl ester signal at 6 Chapter 2 Results and Discussion 60

3.65. Meanwhile, the 13C n.m.r. spectrum clearly showed a signal at o 211.3, indicative of a ketone carbonyl carbon but no ester carbonyl carbon. Hence, the third substance was assigned structure 138, the result of the desired Michael addition but formation of a butyl ketone in place of the methyl ester group.

It was not clear if ketone formation had taken place before or after the conjugate addition.

Repetition of the reaction, using an exactly stoichiomeric equivalent of n-BuLi, gave a similar result, including the formation of ketone 138 as a minor pro.duct. This implied that ketone formation might have taken place before the conjugate addition, in which case the expelled methoxide ion might have served as a base. It is interesting that this side reaction was not observed in the reaction of hydrazone 134 and methyl crotonate (Scheme 36); the much improved yield of the Michael addition process in the reaction from hydrazone 117 was reassuring.

Michael addition of the anion from hydrazone 117 with lactone 30 was next examined (Scheme 36) under the conditions established in the model system. The addition gave the desired product, lactone 139, in an even more attractive 90% yield, but as a 9:79: 12 mixture of three diastereomers 139a-c, respectively. There appeared to be none of the corresponding ketone formed in this case. The 1H and 13C n.m.r. spectra showed well-defined peaks for the three diastereomers. In particular, the 1H n.m.r.

spectrum presented three sets of doublets at o6.47 (JS.I Hz), 6.52 (J 5.1 Hz) and 6.56

(J 4.6 Hz), corresponding to the aldimine protons for isomers 139a, b and c,

respectively, and three sets of doublets at o0.96 (J 6.2 Hz), 1.41 (J 6.2 Hz) and 1.43 (J Chapter 2 Results and Discussion 61

6.2 Hz) from the methyl group at position 5 of isomers 139a, 139b and 139c, respectively.

::rNNMe2 H,, H tJle . ~ ··s 0 1

OMe OMe OMe

139a 139b 139c

Column chromatography on silica gel, eluting with ethyl acetate and light petroleum (1:1) afforded one fraction as a mixture of isomers a and b and another fraction containing isomer c. The mixture of two isomers was subjected to preparative t.l.c. on silica gel to give the major isomer 139b as a pale yellow oil in 65% yield.

Elemental analysis and the appearance in the mass spectrum of a moderately intense molecular ion at mlz 304, were supportive of the expected molecular formula

C17H24N20 3• Distinctive fragment ions appeared at m/z 273 and 205, due to loss of a methoxy radical and the lactone portion from the parent peak, respectively. In the infrared spectrum, lactone 139b showed a strong carbonyl absorption at 1760 cm· 1•

Structure elucidation of isomer 139b was aided by HSQC, HMBC and NOESY n.m.r. experiments, some of the results of which are collected in Tables 4 and 5. The methylene protons next to the lactone carbonyl group were evident as doublets of doublets, at 6

2.13 (dd, J 18.5, 5.6 Hz) and 2.44 (dd, J 18.5, 9.2 Hz), respectively. Meanwhile the 5- methyl protons resonated as a doublet at 6 1.42 (d, J 6.7 Hz) and their neighbour at H5 was evident as a doublet of quartets at 4.64 ( dq, J 4.1, 6. 7 Hz). The two remaining important signals were a complex multiplet at 6 2.89, assigned to H4, and a doublet of doublets at 6 3.92 due to the benzylic methine proton, HI'. Chapter 2 Results and Discussion 62

Table 4. Selected 1H and 13C n.m.r. correlation from 139b through

HSQC and HMBC experiments

Carbon 13C (0) Correlations (0) HSQC HMBC 5-CH3 21.2 1.43 2.89

C3 32.6 2.13, 2.44 3.92

C4 44.6 2.89 1.43, 2.13, 2.44, 3.92, 6.52

Cl' 46.6 3.92 2.13, 2.44, 3.88 (w), 4.64 (w), 6.52

C5 81.5 4.64 1.43, 2.13, 2.44, 3.92

A NOESY experiment indicated the presence of an nOe cross peak between the

5-methyl signal (o 1.43) and the H4 signal at 6 2.89, as well as to 6 4.64 (H5), but no correlations between the signals at 6 2.89 (H4) and 4.64 (H5). These results suggested that the two protons at C4 and C5 were trans. Moreover, a strong cross peak from the signal at 6 3.92 (HI') to the signal at 6 4.64 (H5), but no corresponding correlation to the methyl signal (6 1.43) revealed that H5 and the aryl substituent were cis. A NOESY correlation between the H4 signal (6 2.89) and one half of the methylene signal at 6 2.44 also indicated that the latter signal arose from Ha3. Therefore the relative stereochemistry at C4 and C5 of the major isomer from this reaction was confirmed as that in 139b or 139c. Chapter 2 Results and Discussion 63

H 4.64

139b

Table 5. NOESY experiment on 139b

Proton o NOESY

1.43 2.89, 4.64

2.13 2.44, 2.89 (w), 3.92 (w)

2.44 2.13, 2.89 (w)

H4 2.89 1.43, 2.44, 3.92 (w)

HI' 3.92 2. I 3 (w), 2.89 (w), 4.36, 4.58, 4.64, 6.52

H5 4.64 1.43, 3.92

Surprisingly, when the phenoxyhydrazone 121 was treated with I.I mole equivalents of n-BuLi and Cul as described for the methoxy analogue 117, and lactone -k, ~ '~ 30 was added to the mixture, conjugate reaction of the imine anion from 121£ did not occur. Instead, the unreacted phenoxyhydrazone 121 was the only material recovered, and then in a quantitative yield (Scheme 38). A reddish solution had been observed when hydrazone 117 was added to the n-BuLi solution, and this was taken as an indication of anion formation. However, the red colour was not produced in the case of hydrazone

121. The lack of colour in solutions of hydrazone 121 and n-butyllithium was reproducible but the reason for it was not clear. One possibility was that the amount of base was insufficient. Hence the hydrazone 121 was treated with 2.2 mole equivalents Chapter 2 Results and Discussion 64

of n-BuLi whereupon a red solution was observed. The mixture was then quenched with deuterium oxide. Workup returned the hydrazone 121 as the sole product (63%). 1H

N.m.r. spectroscopic analysis showed an 18% decrease in the integral of the signal at 8

3.68, indicating that some deuteration had occurred at 8 3.68 (H2). Meanwhile, the mass spectrum showed that the recovered hydrazone 121 gave a significantly more intense ion at m/z 269 (10%) and a smaller molecular ion at 268 (6%). These results were supportive of only partial deuteration and therefore partial formation of the desired anionic intermediate, even in the presence of excess base. Hence, in the absence of carbanion formation, the attempted Michael addition of hydrazone 121 with lactone 30 probably led to destruction of the lactone through its reaction with the butyllithium.

Thus no further studies on this reaction have been carried out.

~NNMe2 i,i~ ~ OPh

121 140

Reagents: i. BuLi, Cul, THF, -78°C; ii. butenolide 30

Scheme 38

An inference that could be drawn from these Michael addition reaction results was that the methoxy group assisted in the deprotonation step or stabilized the hydrazone anion. In addition, it was concluded that the lactone 30 was a better Michael acceptor than was methyl crotonate. Chapter 2 Results and Discussion 65

2.5.3 Reactions of arylmethyl bromide 141 with methyl crotonate/lactone 30

Preliminary experiments revealed that the ring closure of intermediates like hydrazonolactone 139 would not be straightforward. One complicating factor was the presence of secondary functionality such as the hydrazono group, that was not necessary for cyclisation but might undergo competing reactions. Hence, a simpler model was sought. Compound 142 was considered ideal for this purpose since it had the necessary elements for the cyclisation and retained stereochemical features that might also play a part in any successful reaction. Hence, attempts were made to synthesise the compound. Retro-synthetic analysis of compound 142 revealed an approach (Scheme

39) that was similar to that developed in the previous section.

Me CH Me ~Br + ~/ ~OMe OMe O

142 141 30 n C(r OMe

143

Scheme 39

The raw material for the forward synthesis, arylbromide 143, was readily

prepared in 99% yield from commercially available 2-bromomethylphenyl bromide 144

through solvolysis in MeOH (Scheme 40). The identity of the product as compound

143 was confirmed through the appearance in its mass spectrum of isotopic molecular Chapter 2 Results and Discussion 66

ions at m/z 202 and 200, and a base peak at mlz 121 corresponding to loss of a bromine radical from the molecular ion. The n.m.r. spectra also showed a three-proton singlet at o 3.47 and a carbon signal at o 58.6 corresponding to the newly introduced methoxy group.

The 2-(methoxymethyl)bromobenzene 143 was next treated with magnesium turnings in anhydrous tetrahydrofuran to afford the Grignard reagent. The reagent was treated with acetaldehyde at -5°C to give the desired product 145 (Scheme 40) in 61% yield after Kugelrohr distillation. The formation of compound 145 was confirmed by elemental analysis, mass spectrometry and 1H and 13C n.m.r. spectroscopy. The molecular ion was absent in the mass spectrum, but an abundant fragment ion at m/z 151 corresponded to loss of a methyl group from the molecular ion. Elemental analysis was consistent with the molecular formula C10H140 2• The substance showed a broad absorption at 3200 cm· 1 in the infrared spectrum, indicative of OH stretching.

Moreover, the 1H n.m.r. spectrum revealed a new three-proton doublet at o 1.51 (J 6.7

Hz) and one-proton quartet at o 5.10 (J 6.7 Hz) that corresponded to the newly

introduced methyl and methine protons, respectively. It was also interesting to note that

the methylene protons on the methoxymethyl group were now non-equivalent (o 4.42

and 4.61, J 11.3 Hz) strongly indicative of the presence of a new stereogenic centre

within the molecule. Thus the data were fully supportive of structure 145. Chapter 2 Results and Discussion 67

Me Br ~Br OC ii, or iii ~OH Br ~OMe ~OMe

144 143 145

~Br ~OMe

141

Reagents: !· MeOH, r~flux 24 h; ii. Mg, THF, (HCHO)n; iii. Mg, THF, CH3CHO IV. HBr (481/o), r.t.,, /3,N MfSOt1 / c.11,._C(,._.

Scheme 40

With the alcohol 145 in hand, its transfonnation into bromide 141 was pursued

(Scheme 40). Treatment with hydrobromic acid under phase-transfer conditions at room temperature gave the desired product 141 in 81 % yield. The mass spectrum showed the expected isotopic molecular ions at m/z 230 and 228, and a fragment ion at m/z 149 corresponding to loss of bromine. The infrared spectrum now lacked the hydroxyl absorption, while the 1H n.m.r. spectrum retained a mutually coupled methyl doublet at o 2.08 (J 6.7 Hz) and methine quartet at 5.62 (J 6.7 Hz), required of the bromoethyl substituent. The methine signal was slightly higher in chemical shift than the corresponding signal in compound 145, supportive of the change from hydroxyl to bromide substituent.

At this stage, there were several options by which the benzyl bromide 141 could be used as a Michael donor. Attempted Michael reactions of benzyl bromide 141 with Chapter 2 Results and Discussion 68

lactone 30 and with methyl crotonate through organometallic and radical intermediates

were investigated and have been discussed in tum below. ln1)m.mi- Initially, a ~-lithium exchange was attempted in order to prepare a suitable organometallic nucleophile. Bromide 141 was treated with n-BuLi in dry THF at -78°C for 30 min and then lactone 30 was added (Scheme 41 ). However, workup did not yield the desired Michael adduct 142. Instead, chromatography of the product gave a 65:35 mixture of meso and racemic dimer 146 and a cross-coupled product 147. Dimer 146 must have arisen through Wurtz coupling, 105 a common side reaction~!rganometallic reagents from benzylic halides. In contrast, the cross-coupled product 147 was probably derived from halogen-metal exchange, but followE?ct by an unwanted nucleophilic substitution involving bromobutane.

~Br + ~OMe Meo OMe

141 146 147

Reagents: i. n-Buli, THF, 30

Scheme 41

Benzylic halides are particularly susceptible to coupling side-reactions, and although there have been many reports of the preparation of benzylic magnesium

halides, the yields are variable and conditions critical. For example, the preparation of

the Grignard reagent from 1,2-bis-(halomethyl)benzene requires106 the use of the

dichloride rather than the dibromide (from which the Grignard reagent decomposes), the

use of THF rather than diethyl ether or other solvent, high dilution (0.075 M) to Chapter 2 Results and Discussion 69

discourage the intermolecular "Wurtz" coupling reaction, 107 freshly distilled halide, and the use of a particular brand (May & Baker Ltd.) of magnesium powder. 108

Although not all these conditions could be satisfied in this thesis work, preparation of the Grignard reagent from the benzylic bromide 141 using Ajax brand magnesium turnings in THF at low concentration (0.066 M), followed by quenching with methyl crotonate at 0°C was examined. The reaction gave only dimer 14b and the conjugate addition product could not be detected.

It has been reported 109• 111 that allylation reactions of aldehydes and ketones occur smoothly with indium powder in aqueous media. The corresponding Michael reaction has not been described but the apparent lack of allyl coupling was attractive.

Therefore, bromide 141 was treated with a mixture of methyl crotonate and indium powder in water at room temperature for 18 hours (Scheme 42). Regrettably, after normal work-up, the reaction gave benzylic alcohol 14S, which was identified by comparison of its 1H n.m.r. spectroscopic data with those of an authentic sample.

Me Me

~Br ~OH ~OMe ~OMe

141 145

Reagents: i. In, H2o, methyl crotonate

Scheme 42 Chapter 2 Results and Discussion 70

A possibility for future study was the use of non-nucleophilic organic solvents in place of water, but this option was not followed. Instead a free radical based conjugate addition pathway was attempted. .J ~ (;i()Ll),ia,tf A mixture of benzyl bromide 141,~ancf"tributyltin hydride in dry benzene containing AIBN was refluxed for 4 hours. Unfortunately, aqueous workup again gave the benzylic alcohol 145 in 83% yield.

It was disappointing that lactone 142 had not yielded to synthesis, despite these varied efforts. However, it was designed merely as a model for ring closure studies. It was therefore decided to return to the fully functional system and to examine procedures for its ring closure. Before embarking on ring closure itself, a series of modified lactone precursors were prepared.

2.6 Preparation of Advanced Intermediates

Two fundamental strategies were adopted in preparing advanced intermediates that might be suitable for a ring closure. In the first, the electron-withdrawing hydrazono group was replaced by more benign and more relevant groups in terms of the overall synthesis. The second approach was convert the methoxy group into a better leaving group, or a more suitable electrophilic group.

2.6.1 Preparation of aldehyde 148 through solvolysis of hydrazone 139

In designing replacements for the hydrazono group, it was decided to first convert the hydrazone 139 into the corresponding aldehyde 148 (Scheme 43) and

subsequently its acetal 149. Cleavage of hydrazones can be achieved either by Chapter 2 Results and Discussion 71

oxidation43 or hydrolysis. 112

Limited solubility of the lactone 139 in water, prompted treatment of the lactone

139 in EtOH with a few drops of 30% aqueous sulfuric acid. It was anticipated that the water in the dilute acid would be sufficient for complete hydrolysis. However, reaction at gentle reflux for 5 hours and workup gave only a trace of aldehyde 148. The major product was a new substance that was isolated as a colourless oil in 81 % yield. The new substance gave a trace fragmentation ion at mlz 335 (M-1) in its mass spectrum and microanalyticalffiiat"\result,£1 were consistent with molecular formula C 19H280 5• Since nitrogen was absent, there was strong evidence that cleavage of the hydrazone had occurred, but the appearance of an additional C4H 100 unit indicated that more than hydrolysis to aldehyde 148 had occurred. In keeping with this suggestion, it was noted that there was no expected aldehyde proton signal in the 1H n.m.r. spectrum. Moreover, replacement of the well-separated methylene proton signals (6 2.13 and 2.44) from the lactone portion of hydrazone 139 by more closely spaced signals at 6 2.42 and 2.45 suggested ring opening of the lactone had taken place. Analysis of the formula suggested that two moles of ethanol might have become involved. The appearance in the 1H n.m.r. spectrum of two triplets at 6 1.12 (J 7 .1 Hz, 3H) and 1.16 (J 7 .1 Hz, 3H), a quartet at 6

4.00 (J 7.1 Hz, 2H) and two doublet of quartet signals at 6 3 .31 (J 7.1, 4.6 Hz, 1H) and

6 3.73 (J9.6, 7.1 Hz, lH) added support and revealed that one ethyl group was close to

a stereogenic centre. The high chemical shift of the quartet at 6 4.00 indicated its

involvement in an ester and probably the one obtained through lactone cleavage. This

was supported in an HMBC experiment which showed a correlation between the

carbonyl carbon signal at 6 172.3 and the proton signal at 4.00. Finally, the presence of Chapter 2 Results and Discussion 72

the four aromatic proton signals, an AB quartet and a methoxy signal indicated that the

methoxymethyl phenyl part of the molecule was intact. Thus, the information to this point indicated the presence of fragments A, B and C.

I /~OR + OCH2CH3 + «OMe 0

A B C

More n.m.r. experiments were carried out in order to fully elucidate the structure. An HSQC n.m.r. experiment revealed a correlation between a proton doublet signal at 6 4.97 and a carbon signal at 6 110. 7 (see Table 6). These signals must have been derived from the former aldimine group and were clearly not from the corresponding aldehyde. Their high chemical shifts were however consistent with an acetal group. In an HMBC experiment, the signal at 6 110. 7 showed correlation to a proton doublet of doublet at 6 3.42 (J 9.6, 4.2 Hz), that from the magnitude of spin couplings must have been vicinal to the acetal proton, and the two non-equivalent methylene protons from the ethoxy fragment C. These correlations fixed the position of the acetal to the former aldimne carbon and demonstrated that the group had become at least a monoethoxy acetal. The HMBC experiment, which had been optimised for two or three bond coupling, also showed a correlation between 6 110. 7 and the doublet of quartet signal at 6 4.62, which must have been derived from the former methine proton at position 5 in the Iactone ring of 139. This indicated a close proximity of the groups and supported a structure in which the oxygen of the lactone ring had become involved in acetal formation. The gross structure of the product was therefore deduced to be ester acetal 150 (Scheme 43). Chapter 2 Results and Discussion 73

139 150

Reagents: i. EtOH, H+, reflux, 5 h; ii. THF/H2O, H+, r.t. 3 d.

148

Scheme 43

The stereochemistry at positions 2' and 3' on the tetrahydrofuran ring of structure 150 was assumed to remain the same as the corresponding positions in hydrazone 139. It then remained to assign the relative stereochemistry at positions 4' and 5 '. The magnitude of spin coupling constants are much less reliable as an indication of stereochemistry of vicinal protons about five-membered rings than they are about six membered rings. However, as summarized in structure D, the magnitudes of the couplings between vicinal protons attached to the furan ring of 150 were clearly of two types. The large spin coupling (3JH,H 9.6 Hz) between H3' and H4' was similar in magnitude to that between H2' and H3 '. Therefore all three protons were assigned as cis. In contrast, the spin coupling between H4' and H5' was much smaller and therefore probably due to trans protons. Hence the structure of the product was tentatively Chapter 2 Results and Discussion 74

assigned with the stereochemistry as shown in structure 150. This stereochemistry was also taken to reflect the relative configuration in the major isomer of lactone 139, i.e.

139b.

Table 6. Selected data from the HSQC and HMBC experiments on compound 150

Carbon C (o) H {o) HSQC HMBC CO2CH~H3 14.0 1.16 1.16, 4.00

5'-OCH~H3 15.2 1.12 1.12, 3.31, 3.73

2'-CH3 17.5 1.27 1.27, 2.87

C2 33.8 2.42, 2.45 1.27 (w), 2.87 (w), 3.42

C3' 46.5 2.87 1.27, 2.42, 2.45, 3.42, 4.62 (w), 4.98 (w)

C4' 51.3 3.42 2.42, 2.45, 2.87, 4.62, 4.97 (w), 7.36 (w)

CO~H2CH3 60.2 4.00 1.16, 4.00

5'-O~HaHbCH3 63.6 3.31, 3.73 1.12, 4.97

C2' 77.7 4.62 1.27, 2.42, 2.45, 2.87 (w), 4.97

CS' 110.7 4.97 3.31, 3.42, 3.73, 4.62

Cl 172.3 2.42, 2.45, 4.00

The unexpected product 150 clearly arose from participation of the solvent ethanol. The following mechanism was proposed for its formation (Scheme 44). Chapter 2 Results and Discussion 75

139b !

150 +

Scheme 44

In order to avoid the participation of hydroxylic cosolvents in the hydrolysis reaction, lactone 139 was treated in aqueous tetrahydrofuran with 30% aq. H2SO4•

Stirring at room temperature for 3 days gave the desired aldehyde 148 in 52% yield

(Scheme 43).

It was noted that although the starting lactone 139b was used as a single isomer, aldehyde 148 was obtained as a 63:37 mixture of diastereomers as indicated by the presence of two aldehydic 1H n.m.r. signals at o 9.53 and 9.58 and their integration, respectively. The diastereomers were also evident from the appearance of two doublet signals at o0.96 (J 6.2 Hz) and 1.45 (J 6.7 Hz), supposedly due to the 5-methyl group.

The two isomers were not separated but characterised as the mixture. The substance gave mass spectral fragment ions at mlz 261 and 233 through loss of an hydrogen and an aldehyde group, respectively, from the parent ion, and elemental analysis results that Chapter 2 Results and Discussion 76

were consistent with the molecular formula C15H18O4• In the 1H n.m.r. spectrum, the distinctive six-proton singlet at o 2. 74 from the dimethylamino group and doublet at o

6.52 (J 5.1 Hz) from the aldimine proton of the starting material 139b were absent, but the remaining signals resembled those of the hydrazone 139b. The assignment as structure 148 was supported by the 13C n.m.r. spectrum, which resembled that of hydrazone 139.

Stereoisomers were thought to arise from epimerisation at the site next to the aldehyde group but leaving the relative stereochemistry about the lactone ring fixed.

With the successful synthesis of aldehyde 148, it was necessary to block the reactive aldehydic group to prevent side reactions later in the synthesis. Two approaches were planned, one through protection of the aldehyde as its acetal to give compound 149 (Scheme 45), and the other through conversion of the aldehyde portion of the molecule through its alcohol and toluenesulfonate derivatives into a phenylthiomethyl group, as in sulfide 153 (Scheme 46).

Me .::-

OMe

148 149 Reagents: i. ethylene glycol, PPTS, benzene

Scheme 45

2.6.2 Preparation of acetal 149

Although acetals are easily hydrolyzed by acids, they are extremely resistant to cleavage by bases, and incorporation of an acetal in this case would avoid competitive Chapter 2 Results and Discussion 77

deprotonation during base-catalysed cyclisation reactions. Thus, the aldehyde 148 was treated with ethylene glycol in benzene in the presence of pyridinium para toluenesulfonate. Reflux of the mixture for 5 hours gave a 54:46 diastereomeric mixture of inseparable isomers as a pale yellow oil in 83% yield. The substance gave in the mass spectrum, the expected molecular ion at mlz 306, a base peak at mlz 73 due to the dioxolanyl ion, and fragment ions at m/z 274 and 201 corresponding to loss of the methoxy group and then loss of the dioxolanyl group, respectively, from the parent molecule. The microanalytical results confirmed the molecular formula as C17H220 5•

Moreover, the 1H n.m.r. spectrum gave a four-proton multiplet (6 3.70-3.80) corresponding to the newly introduced dioxolanyl group. These data confirmed that the desired acetalation had taken place to give acetal lactone 149, but assignment of the relative configuration of the diastereomers was only possible after more detailed n.m.r. spectroscopic analysis.

The most clearly defined signals in the proton spectrum of the isomeric mixture were two sets of doublets at 6 0.84 and 1.49 and 5.00 and 5.05. These were assigned on the basis of their chemical shift and multiplicity to the 5-methyl group and H2", respectively. Analysis of chemical shift data and spin coupling values then permitted assignment of the remaining proton signals (see Table 7 for selected data), which were fully supportive of the proposed structure 149, excluding stereochemistry. A series of

HSQC and HMBC experimentsken allowed assignment of the 13C n.m.r. spectrum. Chapter 2 Results and Discussion 78

Table 7. Selected 1H n.m.r. data for compoundt149a and 149b

Proton 8 149a 8 149b

5-CH3 0.84 (d, J 6.2 Hz) 1.49 (d, J 6.2 Hz)

H113 2.75 (dd, J 17.4, 4.1 Hz) 2.19 (dd,J 18.5, 7.7 Hz)

Ha3 2.78 (dd, J 17.4, ea 5 Hz) 2.40 (dd, J 18.5, 9.2 Hz)

H4 2.71 (dddd, J9.8, 9.1, 7.6, 5.7 Hz) 2.78 (dddd, J9.8, 8.9, 8.9, 5.7 Hz)

H5 4.33 (dq, J 6.2, 6.2 Hz) 4.76 (dq, J 6.2, 6.2 Hz)

HI' 3.38 (dd, J ea 10, 3.6 Hz) 3.48 (dd, J9.8, 4.0 Hz)

H2" 5.00 (d, J3.6 Hz) 5.05 (d, J 4.0 Hz)

The issue of stereochemistry was addressed through a NOESY experiment on acetal 149. Selected data from this experiment are shown in Table 8. Irradiation of the three proton doublet at 8 0.84, due to the 5-methyl, gave correlations with signals at

2.71 (H4) and 4.33 (H5), but there was no correlation between 8 2.71 and 4.33. The former indicated a cis relationship between H4 and the 5-methyl group and the fact that there was no correlation between the signals for H4 and H5 added support. Thus, in this isomer the expected relative stereochemistry at C4 and C5 was evident. Irradiation of the H4 signal at 6 2.71, gave the expected correlation with the signal at 0.84 (5-CH3), but also another at 2.75 (H3). This enabled assignment of the latter signal to Ha3. There was also a weak correlation between the signal at 8 2.71 and that at 5.00 (H2"), but it was difficult to draw conclusions from this observation. Irradiation ofH5 at 6 4.33 gave strong correlations with signals at 0.84 (5-Me) and 3.38 (HI') again supportive of a cis relationship between the H5 and the group containing HI'. Curiously, there was no Chapter 2 Results and Discussion 79

correlation between the signals at 6 2.71 (H4) and 3.38 (HI'). Since the protons at these positions were in a vicinal relationship, the lack of an nOe correlation might have indicated the presence of a preferred conformation (e.g. E) in which the protons were well separated. Thus, the relative stereochemistry at two of the stereogenic centres in isomer 149a were defined. Isomer 149b gave similar correlations between the signals for

5-Me (6 1.49) and H4 (2.78) as in isomer 149a (Table 8), but there was also a weak correlation between the signals at 6 2. 78 (H4) and 3 .48 (HI'). Overall then, the difference between the two isomers must have been their configuration at C 1'.

H~H0- 0 Me A B H

149a 149b E

Figure 2

Table 8. Selected NOESY experiment on 149a and 149b

Proton 6149a NOE correlations 6149b NOE correlations

5-CH3 0.84 2.71, 4.33 1.49 2.78, 4.76, 5.05 (w)

H4 2.71 0.84, 2.84, 5.00 (w), 2.78 1.49, 2.20, 2.39, 5.05, 7.35 3.48 (w), 7.25 H5 4.33 0.84, 3.38 4.76 1.49, 2. 78 (w), 3.48, 5.05

HI' 3.38 2.80, 4.27, 4.33, 2.84 3.48 2.20, 4.24, 4.65, 2.78, (w), 4.64, 5.00 4. 76, 5.05 (w) H2" 5.00 2.80, 3.38, 2.84 (w), 5.05 1.49 (w), 2.78, 3.48, 7.35 3.73-3.85, 4.76, 7.25

Close inspection of Table 8 revealed that the signals at 6 1.49 (5-Me), 2.78 (H4) and 4. 76 (H5) from 149b all correlated with the signal at 6 5.05 (H2"). This pattern was Chapter 2 Results and Discussion 80

not evident in isomer 149a and was taken to indicate a structure such as E in which the group A was the acetal (Figure 2). The earlier isomer 149a was therefore deduced have the opposite configuration at Cl' and therefore the structure in Figure 2.

of 2.6.3 Preparation~ulfide 153

An alternative to acetal formation as a means of blocking the reactivity of the aldehyde 148 was to convert it through the corresponding alcohol 151 and p toluenesulfonate 152 into a phenylsulfide derivative (Scheme 46).

OH SPh _f:Ae ~Me .-o~ ~: I ! .-o 0~

OMe OMe OMe 0 OMe 0

148 151 152 153

Reagents: i. NaBH4, EtOH, r.t.; ii. TsCI, py, 0-4°C; iii. PhSH, KOBu\ DMSO, r.t.

Scheme 46

It was important to retain the lactone carbonyl group at this stage. Selective reduction of the aldehyde 148 was achieved using a small excess (2.4 mol equiv) of sodium borohydride in absolute ethanol. The reaction of a 63 :3 7 diastereomeric mixture of aldehyde 148 gave an 89:11 mixture of diastereomeric alcohols. The mass spectrum lacked the expected molecular ion at mlz 264, but contained a minor signal at mlz 214 and major fragment ions at m/z 202, 159 and 130 (base peak). The high resolution electrospray mass spectrum gave an ion at m/z 287.1256 that agreed with a sodio molecular formula C15H200 4Na. The infrared spectrum clearly showed the presence of Chapter 2 Results and Discussion 81

an hydroxyl function through the appearance of a strong absorption at 3450 cm· 1• The major isomer was isolated by preparative thin layer chromatography in 74% overall yield. In its 1H and 13C n.m.r. spectra, the former aldehyde 148 signals at o 9.58 and

197.8 were replaced with a new set of methylene signals at o (1H) 3.70 and 3.87, and

(1 3C) 65.5, entirely consistent with the gross structure of alcohol 151. The complete assignment of n.m.r. spectroscopic signals from compound 151 were made by comparison with data from aldehyde 148 n.m.r. spectra.

Following the method of Hart,54•56 alcohol 151 was converted into its tosylate

152 by treatment with freshly recrystallized p-toluenesulfonyl chloride in pyridine at 0-

40C for 3 days. It was essential that pure reagent was used in this step to achieve complete reaction. Workup gave a 92:8 mixture of diastereomers, a small enrichment compared to the 89: 11 diastereomeric alcohol 151 mixture that was used in the process.

The major isomer was separated by preparative t.l.c. on silica gel as a pale yellow oil in

80% yield. The p-toluenesufonate 152 gave in its mass spectrum a very weak ion at mlz

419 (M +I), no molecular ion, and a base peak at m/z 91, corresponding to a methylphenyl fragment. The high resolution electrospray mass spectrum gave an ion at mlz 441.1343 that was in agreement with a sodio-molecular formula C22H2606Na.

Infrared absorptions at 1350 and 1170 cm·1 were supportive of asymmetric and symmetric S=O stretching, respectively, while OH stretching at 3450-3360 cm·1 was absent. The 1H n.m.r. spectrum contained a singlet at o 2.41, corresponding to the methyl group of the p-toluenesulfonyl group, and an eight proton multiplet in the aromatic region, consistent with incorporation of the additional aromatic ring. The methylene proton signals of the hydroxymethyl group in alcohol 151 remained Chapter 2 Results and Discussion 82

diastereotopic but were also shifted to 6 4.10 and 4.15 in the new derivative 152, revealing that the group was adjacent to a more electronegative group. Thus, the structure of tosylate 152 was confirmed.

The tosylate 152 could serve as a useful intermediate to a wide range of new derivatives, but was converted in this instance into the phenylsulfide 153 by stirring it with thiophenol in the presence of potassium tert-butoxide in dry DMSO. Reaction at room temperature for 1 hour gave what appeared to be a 93:7 mixture of diastereomers as a pale brown gum in 82% yield. Microanalysis results were consistent with the formula C21 H240 3S, and the expected molecular ion at mlz 356 and a base peak at m/z

123 due to the +CH2SPh ion were visible in the electron impact mass spectrum. The only evidence for diastereomers was found in the 1H n.m.r. spectrum where there was a second doublet signal at 6 0.83 in addition to the major doublet at 1.49, due to the 5- methyl group. Duplicate carbon signals were not evident in the 13C n.m.r. spectrum. The two sets of doublet of doublet signals due to the p-toluenesulfonyloxymethylene were shifted from 6 4.10 and 4.15 in compound 152 to higher field positions at 6 3.28 and

3.38 in the new product. This was consistent with displacement of the p toluenesulfonyloxy group by the less electronegative phenylsulfide group. The methyl singlet and the characteristic AA'XX' pattern of aromatic proton signals from the p toluenesulfonyl group was also replaced by a narrow multiplet 6 7.14-7.34 that was also consistent with introduction of the phenylsulfide group. The remaining proton and carbon-13 n.m.r. signals were readily assigned by comparison of chemical shift values with data from the alcohol 151 and p-toluenesulfonate 152, and with the aid of DEPT experiments. Chapter 2 Results and Discussion 83

There were eight carbon signals in the high field (10-90 ppm) region of the

spectrum of 152, corresponding to the non-aromatic carbons in the molecule. One signal at o 47.4, due to C 1', was very broad and of relatively low intensity compared to the others. A similar observation was made for a related sulfide in later work and the cause will be discussed in page 94.

Thus, the structure of phenylsulfide 153 was confirmed, and the molecule was seen as a key potential precursor for cyclisation studies.

2.6.4 Reactions of compounds 137 and 138 with hydrogen,zromide

Numerous attempts to achieve tricycle formation from sulfide 153 by intramolecular alkylation of its lactone were unsuccessful (see Section 2. 7). A perceived problem with the reaction was the poor leaving ability of the benzylic methoxy group.

The group had reluctantly been introduced early in the synthesis because of the ease of displacement of a bromine atom from the same position. Hence, efforts were made to convert the reluctant methoxy group into a better leaving group, such as a bromide.

There were a number of reagents available for the cleavage of dialkyl ethers, for example,

100% sulfuric acid and other concentrated strong acids, 113 such as constant boiling HI and HBr. 114

In order to conserve stocks of the valuable sufide 153, a number of model studies were carried out using more readily available intermediates. As described in Section

2.4.2, cleavage of isochroman 110, a cyclic benzylic ether, to bromoalcohol 79, was achieved by treatment with hydrogen bromide gas. Cleavage of the benzylic C-0 bond of the advanced intermediates, such atompounds 139 and 153, was expected since this Chapter 2 Results and Discussion 84

outcome would mimic the behaviour of the isochroman and result from formation of the more stable cationic intermediate. It was recognised that another pathway was possible and that cleavage of the ether might also result from demethylation by displacement of methyl bromide by an SN2 process (Scheme 47).

HBr ('yx HBr ('yx CH30H + CH3Br SN1 SN2 ~O-CH3 ~OH

Scheme 47

Thus a solution of ketone 138 in dichloromethane was saturated with hydrogen bromide and the resulting solution kept in a sealed flask for 17 hours at room temperature (Scheme 48).

Aqueous workup gave a mixture of substances from which one pure substance was isolated. The substance gave a weak ion in the mass spectrum, at m/z 241, with no isotopic bromine peak but major fragment ions occurring at m/z 198, 169, 155 and 149

(base peak). A high resolution electrospray mass spectrum gave an ion at mlz 395.1668 which was consistent with the sodio-molecular formula C1sH24O2Na. The infrared spectrum showed a decrease in the carbonyl absorption frequency to 1660 cm·1, which suggested that the carbonyl group was now highly conjugated. The 1H n.m.r. spectrum clearly showed three-proton doublet and triplet signals (o 0.89, J 5.2 Hz and 0.90, J 7.2

Hz, respectively), a two-proton triplet of quartets (o 1.44) and a four-proton multiplet

(o 2.12-2.31) that were thought to have arisen from a methyl group on a secondary sp3 carbon and an isolated propyl group on an sp2 carbon, respectively. Moreover, a quaternary carbon-13 signal at o 199. 7 confirmed retention of a ketone group. Chapter 2 Results and Discussion 85

It was noted that the proton chemical shifts and spin splitting patterns from the methoxymethyl group in the new substance remained the same as in the n.m.r. spectrum of the starting material 138. However, a new one-proton signal at 6 6.49 ( dd, J 2.3, I. I

Hz) that correlated with a carbon signal at 6 148.6 in a heteronuclear multiple quantum coherence (HMQC) experiment (Table 9), indicated the presence of a trisubstituted alkene group in the molecule in which the proton was deshielded. DEPT n.m.r. experiments confirmed the methine nature of the olefinic carbon signal at 6 148.6 and r revealed the presence of a quaternary olefinic carbon Q42.1) and two methylene carbons er ~ 1.6 and 31.1 ). The most logical structure for the new substance was that of the a,~- unsaturated ketone 155. Its formation could be explained through cyclocondensation of the keto hydrazone 138 according to one of the pathways outlined in Scheme 49.

HB~)9) ff • Bu"

138 154

IHBr (g)

155

Scheme 48

The relative stereochemistry at positions 4 and 5 in structure 155 was not Chapter 2 Results and Discussion 86

addressed in the early discussion, but was instructive. The signal for H4 appeared as a doublet of narrow multiplets in which the magnitud:fdoublet coupling was 7.5 Hz. The coupling from H4 to the olefinic proton H3 was measured through the H3 signal as 2.3

Hz. Therefore, the doublet value of7.5 Hz was due to H4-H5 coupling. This was most consistent with a trans diaxal arrangement of protons and therefore trans stereochemistry at C4 and CS.

Table 9. Selected HMQC and HMBC n.m.r. data from ketone 155

Carbon Observed 13C signal HMQC correlation HMBC correlation

CH2CHiCH3 13.8 0.90 1.45

5-CH3 19.6 0.89 2.31, 2.59

CHiCH2CH3 21.7 1.45 0.90, 1.45, 2.30

CH2CH2CH3 31.2 2.30 0.90, 1.45, 6.49

CS 38.5 2.31 0.89, 6.49

C4 45.8 3.67 0.89, 2.18, 2.59, 7.14

C6 46.2 2,18, 2.59 0.89

C2 142.1 4.43, 4.61, 6.49 (w), 7.21

C3 148.6 6.49 2.30

Cl 199.7 2.18, 2.59, 2.31, 6.49 Chapter 2 Results and Discussion 87

Me2NNH 0 HBr -NH2NMe2

7 0 138

IHBr/H 20

0~ 155 ·H2O

~Me OMe

Scheme 49

In retrospect, inspection of Scheme 49 revealed that the stereochemistry must have arisen from the syn stereoisomer of ketone 138.

The presence of a ketone group in the keto hydrazone 138 had clearly complicated the benzyl ether cleavage reaction by providing an alternative reaction pathway. The pathway might or might not have involved prior cleavage of the hydrazone group, as shown in Scheme 49. Thus the closer model substance, ester 137 was also saturated with gaseous hydrogen bromide in CH2Cl2 as described above for ketone 138. Normal aqueous workup was expected to give bromo aldehyde 156 through concomitant hydrolysis of the hydrazone (Scheme 50).

In practice, the reaction gave a new substance that was isolated after chromatography as a pale yellow gum in 46% yield, but it was again not the desired product 156. The substance gave a molecular ion at m/z 232 in its electron impact mass Chapter 2 Results and Discussion 88

spectrum, and an ion at mlz 255.0989 in its high resolution electrospray mass spectrum that was consistent with a sodio-molecular formula C14H 16O3Na. The base peak in the electron impact spectrum appeared at mlz 131, which corresponded to loss of a methyl butanoate fragment from the molecular ion. Also, it was noticed that the difference in the observed molecular weight and that of the desired substance 156 was 80 atomic mass units. This information led to consideration of a structure related to 156 but one that had lost the elements of hydrogen bromide. Strangely, there was no evidence in the infrared,

1H or 13C n.m.r. spectra of an aldehyde group. The 1H n.m.r. spectrum showed the presence of a methyl doublet at 6 1.23 (d, J 5.3 Hz), the AB portion of an ABX spin system at 2.32 (dd) and 2.67 (dd), and of a methoxy signal at 3.68 (s). A complex one proton multiplet at 6 3.25-3.37 was consistent with the remaining proton of a 3- substituted methyl butyrate fragment. Hence, 1H n.m.r. spectroscopy supported the prediction from mass spectrometry of retention of the butyrate ester portion of the starting hydrazone ester 137. A new doublet at 6 6.49 with a small spin coupling constant (J 1.1 Hz) indicated the presence of a relatively isolated trisubstituted double bond, a feature supported in the 13C n.m.r. spectrum by two new carbon signals at 119.9

(quaternary carbon) and 142.0 (CH), yet there was no longer an aldirrkie proton signal nor a methine signal for a proton next to the former imine group. There did appear an

AB quartet at 6 4.90, 4.96 (J 12.5 Hz), that supported the presence of non-equivalent methylene protons derived from the former methoxymethyl group, yet the methoxy signal was absent. Hence the benzo-pyran structure 157 was proposed for the substance. Chapter 2 Results and Discussion 89

OMe

137 156

Ii. ii

~OMe

157

Scheme 50

This unexpected result could be explained by initial hydrolysis of the hydrazone and then either benzyl ether cleavage and intramolecular cyclisation of the resulting bromoaldehyde (Scheme 51, path A), or methyl ether cleavage and dehydrative cyclisation of the intermediate hydroxy aldehyde (path B). However, since every effort was made to exclude moisture from the mixture during reaction, hydrolysis to the aldehyde 156 might not have been involved. An alternative was demethylation of the methoxymethyl group followed by cyclisation of the hydroxy hydrazone (path C). Chapter 2 Results and Discussion 90

,:,-NNMe2

HBr (A) OMe -H2O OMe - OMe OMe

137

OMe OMe

OH OH 157

j 1l

~OMe ~OMe 0 , 0 HNMe2 H

Scheme 51

The possibility that demethylation had occurred rather than debenzylation, prompted a search for alternative approaches to the cyclisation process that might take advantage of the unexpected direction of cleavage.

2.6.5 Preparation of aldehyde 158

Nitronium tetrafluoroborate (NO2BF 4) has been used to oxidatively cleave alkyl methyl ethers to yield aldehydes. 115 The intermediate oxonium ions decompose with loss of nitrous acid and the direction of proton loss is dictated by the stability of the Chapter 2 Results and Discussion 91

incipient cation. In most methyl ethers the methoxycarbenium ion ensues, and the net reaction (with hydrolytic workup) affords formal demethanation (for an example see

Scheme 52). This method would provide alternative intermediates whose use would differ from previous strategies of ring closure by relying upon an aldol-like condensation rather than an alkylation. -

!-HN02 © a-CH,..,0-Me

Scheme 52

Lactone 139 was treated with nitronium tetrafluoroborate in dichloromethane at ice temperature under argon and the mixture then allowed to warm to room temperature

(Scheme 53). An aqueous quench and extractive workup gave a complex mixture of products by analytical thin layer chromatography. 1H N.m.r. spectroscopic analysis revealed three small singlet resonances in the range 6 9.55-9.85, corresponding to various

aldehydic products, and loss of the distinctive peak at 6 2. 74 corresponding to the N,N

dimethylamino protons in the starting lactone 139. The integral of the methoxy singlet at

6 3.38 was also substantially decreased. Attemptsto separate the mixture by preparative

t.l.c. was unsuccessful.

Since the hydrazone group of compound 139 might have reacted in preference to

the methyl ether, a different substrate, namely, acetal 149, was selected for reaction with Chapter 2 Results and Discussion 92

NO2BF4 (Scheme 53). The reaction was performed at a lower temperature (-23°) than before in dry CH2Cl2 for 45 min and then at 0°C for another 1 hour. An aqueous quench at -23°C and extractive workup gave a product mixture that was much less complex in its 1H n.m.r. spectrum than the first. It showed a 72:28 mixture of unreacted acetal 149, indicated by two doublets at 6 5.00 (J 3.8 Hz) and 5.05 (J 3.8 Hz) in the 1H n.m.r. spectrum, and a new substance, indicated by the presence of two unequally intense singlets at 6 9.55 (minor) and 9.61 (major). The mixture was subjected to preparative thin layer chromatography on silica gel using EtOAc-CH2Cl2 (18:82) as solvent to afford impure aldehyde 160 as a pale yellow oil (20%), and acetal 149 (65%). Attempts to further purify the aldehyde 160 by preparative t.I.c. on silica gel led only to gradual loss of material and the process was abandoned.

Scheme 53

. 1 . . h . 116 Captodative radicals are amongst the most use ful rad 1ca s m orgamc synt es1s.

They are known to be relatively stable and comparatively easy to form. In this case, bromination was expected to yield benzylic bromination, next to the methoxy group of several of the substrates. Such intermediates should be hydrolytically unstable and yield Chapter 2 Results and Discussion 93

aldehydes.

It was thought that the hydrazone 139 and the acetal 149 would undergo competing reactions with bromine or its equivalents. Hence, sulfide 153 was treated with a slight excess of N-bromosuccinimide in carbon tetrachloride in the presence of benzoyl peroxide (Scheme 54). The mixture was refluxed under white light for 3 hours then cooled and quenched with water. The soluble organic phase gave in 31 % yield a mixture of the desired aldehyde 158 and unreacted sulfide 153 by 1H n.m.r. spectroscopic analysis. The reaction was repeated for a longer time (6 hours), but with a similar outcome. However, when 2.5 mole equivalents of N-bromosuccinimide was used and the reaction limited to three hours, the product mixture was more complex.

The mixture was subjected to preparative t.l.c. on silica gel to afford aldehyde

158 in 95% purity as a pale yellow gum in 57% yield. The aldehyde 158 could not be separated from traces of unreacted methyl ether 153 even after extensive chromatography because the two compounds were very similar in Rr value.

The substance gave a single isomer and the routine mass spectrum gave a strong molecular ion at mlz 340. A high resolution electrospray mass spectrum gave a sodio molecular ion at mlz 363.1029 which was consistent with a molecular formula

C20H20O3SNa. There were two carbonyl absorptions, at 1760 and 1680 cm· 1 in the infrared spectrum which corresponded to the lactone and aldehyde carbonyl groups, respectively. In many respects the 1H n.m.r. spectrum indicated that the desired conversion had taken place. The distinctive methoxymethyl signals at o 3.24 (s), 4.15

(d) and 4.47 (d) in the sulfide 153 were absent in the spectrum of the new product.

Instead, the n.m.r. spectra gave signals at o (1H) 9.95 (s) and o (13C) 194.S (CH), that Chapter 2 Results and Discussion 94

indicated that the methoxy group had been transformed into an aldehyde. However, notably, integration of the proton signals in the non-aromatic region (0-5 ppm) corresponded to only nine protons. This contrasted with the ten protons expected from the desired product 158. Moreover, the 13C n.m.r. spectrum gave well-defined carbon signals for the aromatic carbons and an aldehydic carbon, but gave very broad and almost undetectable signals in the high field region ( 10-70 ppm) at room temperature.

153 158

Reagents: i. NBS, CCl4, benzoyl peroxide, white light, reflux

Scheme 54

This prompted a more thorough investigation of the n.m.r. spectra at 55°C at a field strength of 500 MHz for the proton spectrum and 125.8 MHz for the carbon-13 spectrum. Under these conditions, the 1H n.m.r. spectrum of a deuterochloroform solution gave a sharpened signal at 6 2.75 and the 13C n.m.r. spectrum also showed five well-defined carbon signals in the high field region (10-70 ppm), with the exception of a broad signal at 39.0. Although this was an improved result, the broad carbon signal at 6

39.0 was still of concern and prevented complete analysis of the spectra. The temperature of the n.m.r. probe was therefore lowered to -50°C, and the spectra recorded. At this temperature, the expected number of proton signals was observed and Chapter 2 Results and Discussion 95

all were well-defined. The 1H n.m.r spectrum clearly showed for the first time a signal at

8 4.54 (ddd) due to HI' and there was a correlation between it and the now sharp signal at 8 39.0 (Cl') in an HSQC experiment at the low temperature. In addition some 1H n.m.r. signals showed minor duplicate signals at-50°C. These were most clearly evident in the aldehyde region where at-50°C signals appeared at 8 9.95 and 9.83 in the integral ratio 28: 1, respectively.

These low temperature experiments revealed signals that were fully supportive of the structure 158. Complete assignment of the major 1H and 13C n.m.r. signals of the compound 158 was then made at -50°C using HSQC, HMBC and NOESY experiments. clo.to.f,YQ)n Table 10. Selected{~OESY experiment at 223K on compound 158

Proton 8 NOESY Correlation

5-Me 1.48 4.48

H113 2.25 2.47, 2.75 (w), 4.54

H0 3 2.47 2.25, 2.75

H4 2.75 2.25, 2.47 (w), 3.20, 3.34 (w), 4.48 (w), 4.54

SCHaHbPh 3.20 2.75. 3.34, 4.48 (w), 4.54

SCHJI~h 3.34 3.20, 4.48 (w), 4.54

H5 4.48 2.75 (w), 3.20 (w), 3.34 (w) 1 1.if

Hl' 4.54 2.25, 2.75, 3.20 (w), 3.34

H6" 7.41 2.25, 2.47 (w), 2.75, 3.20, 3.34 (w)

CHO 9.94 4.54, 7.80

Selected data from the NOESY experiment were collected in Table 10. The Chapter 2 Results and Discussion 96

relative stereochemistry at positions 4 and 5 on the lactone portion remained the same

as the corresponding position in sulfide 153. The correlations between signals at 6 2.47

(H0 3) and 2.75 (H4) were stronger than those between signals at o 2.25 (H 133) and 2.75

(H4), therefore the higher chemical shift H3 signal was assigned as H0 3. This pattern

was found throughout the advanced intermediate compounds, such as in 139, 149 and

153. Irradiation of the doublet of doublet of doublet at 6 4.54, due to HI', gave an nOe

correlation with the signals at 2.75 (H4) that indicated a cis relationship between HI'

and H4. Thus, the relative stereochemistry of compound 158 was confirmed and the

S/p,r~c chfmi,; t-0 ~t position C 1' was found to be retained as in major isomer 139b even after changing the functio{group.

The broad signals from compound 158 at room temperature were attributed to a

slow exchange rate on the n.m.r. timescale between two extreme conformers. This

phenomenon is common with some classes of compounds, such as amides, but is much

less seen in compounds with single bonds. The exchange between two unequal

populations of conformations has been discussed. 117 The powerful broadening of C I'

was assumed to indicate that this was the centre of any restricted rotation. While the

Ar-CHO bond might have given rise to rotamers, this should not have affected the HI'

or Cl' nuclei. It was further noted that the chemical shift of HI' at 6 4.54 was high for a

benzylic proton. Thus, it was inferred that HI' resided in the same plane as the aromatic

ring in the preferred conformation. The chemical shifts of signals of HI ' ( ddd) in the two

conformers [6 4.54 (major) and 4.65 (minor)] were clearly influenced to a major extent

(.L\v 25 Hz) by this conformational preference. It was more likely that rotation about the

Ar-C 1' bond was being observed. There were two extreme conformations, A and B, that Chapter 2 Results and Discussion 97

might arise. One could argue that since a strong nOe correlation was observed between the aldehyde proton and H 1' in the major isomer, then the major isomer was conformer

A. However, it was not possible to detect an nOe correlation for the minor isomer aldehyde signal that could confirm its assignment as the alternative, isomer B (Figure 3).

This might have been due to very low intensity of the minor carbon and proton signals from isomer B.

H tv1e /-: .- R = ¼° 0 A B

Figure 3

On balance, the evidence supported the gross structure of 158 and despite the lack of evidence for conformation B, the conformation of preference for the molecule was as shown in structure A. Thus, the structure of aldehyde 158 was confirmed and the molecule provided an alternative intermediate for an aldol-like condensation in the ring closure study.

2. 7 Ring Closure

The remaining task in the synthesis of the tricyclic portion of the himbacine

analogues was ring closure of the central part of the molecule. As has been described, a

range of precursor types had been made accessible by synthesis. The most readily

available included precursors with a methoxymethyl group on the aromatic ring and

various substituents, such as hydrazone, acetal and sulfide groups, on the central part of

the molecule (Figure 4). These substances were therefore examined first for their ability

to cyclise. Though not an ideal leaving group, the methoxy group was thought to be Chapter 2 Results and Discussion 98 sufficiently good to enable cyclisation to occur by an intramolecular alkylation process on the lactone enolate.

139 R = CH=NNMe2 149 R = CH(OCH2CH20) 153 R = CH2SPh

Figure 4

Lithium diisopropylamide is a commonly used, strong, non-nucleophilic base with wide applicability in organic synthesis. 118 Especially important, it has been used for ester and lactone enolate formation and so was the base of first choice for promoting the ring closure. It has been used for alkylation of esters and lactones, 119 and this method was adopted here. Lithium diisopropylamide was prepared in tetrahydrofuran, and lactones 139 and 149 were added to the base at -78°C in separate reactions (Scheme 55).

Unfortunately, these reactions under a variety of conditions (Table 11) afforded unreacted lactone starting materials 139 and 149, but none of the desired cyclisation products. When lactone 139 was treated with 2.0 equivalents of base and the reaction mixture heated to reflux (entry 3) in an attempt to force ring closure, workup gave a substance with broad signals in the 1H n.m.r. spectrum, but there was no indication of the starting material 139 nor of the desired product 161. This indicated that the lactone had either decomposed or was· polymerized under the vigorous conditions. Chapter 2 Results and Discussion 99

r·

Reagents: i. LOA, THF, -78°C

Scheme 55

Table 11. Reactions oflactones 139 and 149 with LDA

Entry Reactant LDA Temp. Time Quenching Yield(%) (mole e9uiv) (OCl (hl rea~ent 139/149 other 1 139 I.I -78-r.t. 1.75 (r.t.) H2O 44

2 139 I.I -78 2.0 163 62

3 139 2.0 -78-reflux 1.5 NH4Cl (reflux) (Sat.) 4 149 1.2 -78 1.0 H2O 80

5 149 1.2 -78-r.t. 20 (r.t.) H2O 46

6 149 2.0 -78-r.t. 1.75 (r.t.) H2O 76

In order to test the reactivity of the enolate, the lactone 139 was also treated in the normal manner with 1.1 mole equivalents of base and the supposed anion quenched with 3,4-dimethoxybenzaldehyde 163 (Table 11, entry 2) but again only unreacted lactone 139 was recovered. This implied that if deprotonation had been successful and the enolate had been formed, then the enolate was very unreactive.

One possible solution was to add a cosolvent, such as tetramethylethylenediamine or hexamethylphosphoric triamide to complex the lithium cation and thereby increase reactivity of the anion. Another remedy was to change the Chapter 2 Results and Discussion 100

metal cation from lithium to a more ionic cation, such as potassium. Potassium hexamethyldisilazide (KHMDS) is another frequently used hindered base that affords anions with more ionic character. When model ester 137 was added to a solution of potassium hexamethyldisilazide (Scheme 56) at - 78°C and the reaction mixture quenched with 163 or tert-butyldimethylsilyl chloride (TBDMSCl) (Table 12, entries 1 and 2), there was recovered unchanged ester 137 in low to moderate yield. None of the desired products 164 or 165 were detected by 1H n.m.r. spectroscopic analysis. The disappointing results from ester 137 were possibly due to the flexibility of the butanoate ester part of the molecule, which might have given an enolate with an unfavourable conformation. Similar reactions were therefore carried out on the corresponding lactone 139.

137 164

OMe Reagents: i.KHMDS, THF/toluene, 163, -78°C OMe OTBDMS ii. KHMDS, THF, TBDMSCI, -78°C

165

Scheme 56 Chapter 2 Results and Discussion 101

Table 12. Reactions of compounds 137, 139 and 149 with KHMDS

Entry Reactant KHMDS Temp. Reaction Quenching Yield(%) (mol equiv) (OC) time, (h) reagent s.m. other I 137 1.5* -78 1.5 39

2 137 1.3 -78-r.t. 2.0 (r.t.) TBDMSCl 50

3 139 I.I -90 1.75 26 (166) 39 (148) 4 139 I.I -90 1.75 71

5 149 1.3 -90 1.5 89

6 149 I.I -90 1.5 TBDMSCl -

7 149 1.4 -90 1.5

* THF:toluene (1:1)

A solution of lactone 139 was added to a solution of potassium hexamethyldisilazide at -90°C for 15 mm and then a solution of 3,4- dimethoxybenzaldehyde 163 was added (Scheme 57, Table 12, entry 3). Aqueous workup and solvent extraction gave a product containing two substances by 1H n.m.r. spectroscopic analysis. The least polar component was isolated in 35% yield by preparative thin layer chromatography. and identified as unreacted lactone 139. The more polar substance was subsequently identified as the desired condensation product

166. It was obtained as a pale yellow amorphous powder (m.p. 62-64°C) in 26% isolated yield. N.m.r. spectroscopic analysis revealed that the substance was an equimolar mixture of two diastereomers. For example, there appeared duplicate sets of singlets at 6 2.71 and 2.72 for the dimethylamino protons. Similarly, the methoxy group on the dimethoxyphenyl substituent appeared as duplicate pair of signals at 6 3.69 and

3.73, and 3.88 and 3.89. In addition, there were a duplicate signals throughout the 13C Chapter 2 Results and Discussion 102

n.m.r. spectrum. The compound gave a trace parent ion at mlz 470 in the mass spectrum

and elemental analysis was consistent with the molecular formula C26H34N20 6• In the

infrared spectrum there appeared a strong hydroxyl group absorption at 3335 cm· 1•

There was also a prominent mass spectrometric fragment ion at m/z 452 due to loss of a water molecule that supported the presence of the hydroxyl group. The 1H n.m.r. spectrum showed loss of the ABX spin system that was characteristic of the unsubstituted lactone ring and prominent in the spectrum of lactone 139. Instead, there appeared two new broad singlets at o 4.68 and 4.71 and a broad singlet at 4.85 that corresponded to the methine proton at C 1 ' in the two diastereomers and its attached hydroxyl group, respectively. The 13C n.m.r. spectrum gave three new mid-range carbon

signals at o 55.6, 55.9 and 74.9 due to C 1' and the carbons of the two methoxy groups

on the benzenoid ring.

i, ii

OMe

139 A = CH=NNMe2 166 A= CH=NNM~ 149 A = CH(OCH2CH20) 167 A = CH(OC~CH20)

Reagents: i. KHMDS, THF, -90°C; ii. 3,4-dimethoxybenzaldehyde 163, -90°C

Scheme 57

Since the mass recovery of the latter product was low, the original aqueous

solution was neutralised with 30% aqueous sulfuric acid, the solution allowed to stir for Chapter 2 Results and Discussion 103

4 hours and then extracted with ethyl acetate. Evaporation of solvent from the extracts gave another substance in 39% yield that was identified as aldehyde 148. The pale yellow gum was identical in all aspects to that obtained from acidic hydrolysis of lactone 139 in aqueous tetrahydrofuran.

Despite the low yield, formation of the aldol product 166 indicated that the lactone enolate from hydrazone 139 had been generated to at least some extent. With this encouraging result, the reaction was repeated, but instead of addition of the aldehyde 163 the mixture was kept at -90°C for about 2 hours in the hope that cyclisation might take place ( entry 4). Unfortunately, extractive workup afforded only unchanged lactone 139 with 71% recovery. Subsequent attempts to quench the intermediate anion with trimethylsilyl chloride and with t-butyldimethylsilyl chloride afforded only unchanged hydrazone lactone 139. Hence reactivity of the enolate anion was very limited.

Ester 137 and lactone 139 both contained a hydrazone group and concern was expressed that deprotonation might have occurred next to this group, in competition with the desired ester and lactone enolate formation, respectively. A different substrate, acetal 149, was therefore added to 1.4 mole equivalents of potassium hexamethyldisilazide solution and after 1.5 hours the mixture was treated with aldehyde

163 (Scheme 57). Disappointingly, workup afforded only a mixture of acetal 149 and aldehyde 163, from which the acetal 149 was recovered by chromatography in 62% yield. None of the condensation product 167 was observed by 1H n.m.r. spectroscopic analysis. Repetition of this reaction under the same conditions gave the same results.

It appeared from this study with compounds 137, 139 and 149 that enolate Chapter 2 Results and Discussion 104

fonnation could only be demonstrated in the case of the hydrazone lactone 139.

Moreover, the yield obtained by trapping the enolate was low and could only be carried out with an external aldehyde. The absence of cyclised product confinned that the methoxy group was a reluctant leaving group.

Boron enolates have also been used for alkylation reactions. Moreover, they have been used in reactions with aldehydes wherein the reactions have been catalysed by

Lewis acids. Such use of Lewis acid (LA) catalysis in the case of boron enolates derived from hydrazono, acetal or phenylsulfide lactones 139, 149 and 153, respectively,

(Scheme 58) was thought to be potentially a much more successful route to cyclisation because the methoxy group would be made a better leaving group in the process.

139 RH Me ~Me RH ~e or 149 Bu2BOTf LA 0 ---- or ::::,,,... ~o 153 (/Pr)2NEt ~ ----CCR LA Q. OBBu2 H O OMe 08Bu2 -(±)-Me

Scheme 58

Boron enolates are nonnally generated in situ and are not isolated. 120 Hence, as a readily available model, ester 137 was treated under relatively standard conditions with

1.1 mole equivalent of di-n-butylboron triflate and diisopropylethylamine in dry dichloromethane at 0°C for 30 minutes. The supposed boron enolate was then quenched with 3,4-dimethoxybenzaldehyde 163 at -78°C (Table 13, entry 1). Workup afforded a brown oil that chromatographed to yield unchanged ester 137 in 80%.

The reaction was repeated under the same conditions but without treatment with Chapter 2 Results and Discussion 105

3,4-dimethoxybenzaldehyde 163, but again workup gave unchanged ester 137 in 60%

yield.

Table 13. Reactions of compounds 137 and 153 with Bu2BOTf/EtN(iPr)2

Entry Reactant Bu2BOTf/ Temp. Reaction Quenching Yield (0t'1) EtN(iPr)i ( e9) (OC) time (h) rea~ent s.m. other 1 137 1.1/1.1 0-+-78 1.5 (r.t.) 163, H2O 80

2 137 1.1/1.1 -78 1.5 (r.t.) H2O 75

3 137 5.0/5.0 -78 21 (r.t.) H2O

4 153 1.3/1.3 0-+-78 1.0 (0°C) H2O 70

5 153 1.3/1.3 0-+ -78 1.0 (0°C) BF3·OEt2 77 H2O 6 153 only 3.0 eq r.t. 3.5 days H2O BF3·OEt2

It was not clear from these results that boron enolate formation had occurred.

Therefore the reaction was repeated using 5 mole equivalents of di-n-butylboron triflate

and Hunig's base, diisopropylethylamine, in dry dichloromethane, and the mixture was

stirred for 21 hours at room temperature. Evaporation of the solvent without aqueous

workup gave none of the desired boron enolate nor of the desired cyclised product by

1H n.m.r. spectroscopic analysis.

The problem of secondary reaction sites with ester 137 and lactones 139 and 149

probably also existed for this reaction. Hence, the reactions of the phenylsulfide lactone

153 were investigated. The sulfide group was thought unlikely to complex with boron trifluoride. Addition of the sulfide 153 to a mixture of 1.3 mole equivalents each of di-n

butylboron triflate and Hunig's base diisopropylethylamine, at -78°C, reaction for 2

hours, then extended reaction without (Table 13, entry 4) or with the addition of Chapter 2 Results and Discussion 106

borontrifluoride etherate (BF3·0Et2) (entry 5) for an hour at -78°C, and another hour at

0°C, gave only unchanged sulfide 153, in both cases.

The sulfide 153 was then treated with 3 mole equivalents of borontrifluoride etherate in dichloromethane at room temperature for 3.5 days. After normal extractive workup with dichloromethane, there was neither significant starting material recovered nor desired product 169 detected by 1H n.m.r. spectroscopic analysis (entry 6).

The lack of reactivity of the methoxy group, even under Lewis acidic conditions was most disappointing. Attention was therefore turned to the aldehydic lactone 158. o.ffu Cyclisation of this intermediate would afford an alcohol 170 that.if,dehydration and then hydrogenation should afford the desired tricycle 171 (Scheme 59).

SPh

------~ i, iii ------~ r ii, iii

158 170 171

Reagents: i. LOA, THF; ii. KHMDS, THF; iii. H20

Scheme 59

The aldehyde 158 was first treated with 1.2 mole equivalents of lithium diisopropylamide in dry tetrahydrofuran at -78°C for 1.5 hours and the anion quenched with water. The reaction gave a brown oil, in 95% mass recovery, which was not the starting material 158 by 1H n.m.r. spectroscopic analysis. Attempts to separate the complex mixture by preparative thin layer chromatography on silica gel were Chapter 2 Results and Discussion 107

unsuccessful, but yielded some evidence that diisopropylamine might have reacted with the aldehyde group of 158 before condensation could occur.

In earlier work it was found that potassium hexamethyldisilazide was a superior base to lithium diisopropylamide. Therefore, aldehyde 158 was treated with potassium hexamethyldisilazide in dry toluene at -90°C for 1.5 h. Aqueous quench and extractive workup gave unchanged aldehyde 158 in 87% yield.

There was no evidence for the amine in this case having reacted with the aldehyde. However, the strong base might have deprotonated at the benzylic site in preference to the lactone because a quinomethide-like anion would have been fonned.

At this point, lack of material and time to prepare more intennediates prevented further attempts to achieve cyclisation and the study was drawn to a close.

2.8 Summary and Future Directions

In this study, though the complete southern portion of the himbacine analogue

64 was not obtained, much background chemistry was investigated. A method for assembling the carbons of the southern portion of aromatic himbacine analogues was developed through Michael addition of a hydrazone anion to butenolide 30. An improved method for the preparation of Michael acceptor lactone 30 was devised that reliably gave the substance in 51 % overall yield. The initially targeted Michael donor 77 was eventually synthesised but was found unsuitable as an intennediate. An alternative

Michael donor 117 was synthesised and found to be a more stable partner in the

Michael reaction. It was found suitable for the preparation of various potential

intennediates. However, cyclisation of these intennediates proved difficult. In Chapter 2 Results and Discussion 108

particular, the ring closure of the advanced intermediates, 139, 149, 153 and 158 returned starting material or led to poor mass recovery.

Based on the knowledge gained from the present studies, future research should be directed towards continuing efforts to cyclise the current set of advanced lactone intermediates, especially 139, 149, 153 and 158, and to examine the replacement of the methoxy leaving group with a more labile substituent.

There is considerable scope for modifying reaction conditions through use of

alternative bases and formation of enolate equivalents that could be pursued in the first

study. Alternatives to the methoxy leaving group include more readily cleaved groups

such as trialkyl silyloxy, and p-methoxybenzyloxy substituents. 109

Chapter 3 Experimental GENERAL

All commercial materials employed in this research were used as received, except where indicated. Melting points were determined on a Mel-Temp or a Kofler hot-stage melting point apparatus and are uncorrected. Infrared spectra were recorded on a Perkin-Elmer 298 spectrophotometer and refer to thin films of oils or paraffin mulls or KBr discs of solids. Microanalyses were carried out by Dr P. H. Pham of the School of Chemistry Microanalytical Laboratory or Dr R. Bergman, Microanalytical Service Unit, Research School of Chemistry, Australian National University. Optical rotations were measured on a Bellingham and Stanley Ltd polarimeter. Routine 1H n.m.r. spectra were recorded at 300 MHz on a Bruker AC300F spectrometer and at 500 MHz on a Bruker DMX500 spectrometer. Unless otherwise stated, data refer to solutions in deuterochloroform with the residual solvent protons as internal reference. N.m.r. multiplicities were designated as singlet (s), doublet (d), triplet (t), quartet (q), quintet (quint), multiplet (m) and broad (br). 13c n.m.r. spectra were measured at 75.6 MHz on the AC300F instrument and at 125.76 MHz on the DMX500 spectrometer. DEPT carbon spectra were recorded on approximately 0.01 M solutions in CDC13 at 300K. Chemical shifts were in parts per million (o) relative to solvent nuclei as the internal reference. C-H correlation experiments were carried out on the Bruker AC300F spectrometer at 300K using the Bruker automation program XHCORR.AU or on the Bruker DMX500 spectrometer using HSQC or HMBC automation programs, H-H correlations were recorded on the same instruments using the program DQF-COSY. NOESY experiments were conducted on non-spinning samples and a floated magnet at 298K on the Bruker DMX500 spectrometer by Mrs H.

E. R. Stender.

Routine mass spectra were recorded at 70 eV with a VG Quattra spectrometer by Dr J. J. Brophy, while high resolution mass spectra were recorded by Dr R. Lidgard of the UNSW Biomedical Mass Spectrometery Unit on a VG Autospec Q instrument and Mr. R. Zhang of the School of Chemistry on a FTICR mass spectrometer. Chapter 3 Experimental 110

3.1 Preparation of Michael acceptor, (5S)-methyl-2(5H)-furanone 30

Preparation of ethyl (2S)-[(1 '-ethoxy)ethoxy]propanoate 72

Preparation of pyridinium p-toluenesulfonate (PPTS): p-Toluenesulfonic acid monohydrate (5.70 g, 30.0 mmol) was added to pyridine (12.l mL, 150.0 mmol) with stirring at r.t. After 20 min, the excess pyridine was removed by rotary evaporation over a water bath at 60°C to afford a slightly hygroscopic colourless solid. Recrystallization from acetone gave pyridinium p-toluenesulfonate as white needles (5.30 g, 71 %) m.p.

120°C (lit. 121 m.p. 120°C).

Reaction: Ethyl vinyl ether (40 mL, 418 mmol) was 2' 1 • I ' 2·/'-0+0 added to a solution of ethyl (S)-lactate (10.0 g, 85 mmol) Me'2'C02CH2CH3 3 1 and PPTS (10.6 g, 42 mmol) in dry CH2Cl2 (150 mL) at 72 0°C with stirring. A drying tube was fitted, the cooling bath was removed, and the mixture was stirred at r.t. for 20 h. The mixture was washed with water (3 x 200 mL), dried over MgSO4, and the solvent removed in vacuo to yield a colourless oil. The residue was Kugelrohr distilled to afford a l : 1 mixture of two diastereoisomers of ethyl

(2S)-[(l '-ethoxy)ethoxy]propanoate 72 as a mobile oil (15.7 g, 97%) b.p. 64-65°C/0.03 mmHg. IR Vmax (film): 2980, 2920 (br), 1740, 1440, 1370, 1265, 1140, 1080, 960, 850 cm·1• 1H n.m.r. 6 (CDCh, 500 MHz): 1.14, 1.16, 2 x t,J7.1 Hz, (H2")3; 1.26, 1.28, 2 x t, J 7.1 Hz, CH 2CH3; 1.33, 1.39, 2 x d, J 6.9 Hz, (H2')3 and (H3)3 (overlapping); 1.36, d,

J 6.9 Hz, (H3)3; 3.44-3.50, m, IH, 3.57-3.63, m, IH, (Hl")i; 4.14-4.24, m, CH2CH3;

4.18, 4.31, q, J 6.9 Hz, H2; 4.66, 4.75, 2 x q, J 5.4 Hz, HI'. 13C n.m.r. 6 (CDC13, 125.8

MHz): 14.17, 14.19, CHiCH3; 15.08, 15.29, C2"; 18.92, 18.96, C3; 19.71, 20.02, C2';

60.19, 60.22, Cl"; 60.80, 61.51, CH2CH3; 69.63, 69.90, C2; 99.26, 99.39, Cl'; 173.41, Chapter 3 Experimental 111

173.66, CO. Mass spectrum: mlz 190 (M\ absent), 175 (M-15, 17%), 145 (36), 115

(14), 103 (30), 101 (43), 87 (11), 73 (100), 55 (18), 45 (82).

N.m.r. spectroscopic assignments were supported by a HSQC experiment in CDC13.

Preparation of (2S)-[(1-ethoxy)ethoxy)propan-1-ol 73

A solution of ester 72 ( 4.00 g, 21.1 mmol) in diethyl 1· I~2' ether (30 mL) was added dropwise to a vigorously stirred 2':/'-0~0= suspension of LiAlH4 (2.0 g, 52.5 mmol) in diethyl ether (100 73 mL). The mixture was refluxed for a further 2 h, and then treated with water (10.0 mL).

The suspension was filtered through a pad of filter aid and the pad washed several times with more ether. The combined filtrate and washings were dried over MgSO4, and evaporated in vacuo at r.t. to yield a mobile oil (3.10 g).The oil was distilled to obtain a

60:40 diastereomeric mixture of (2S)-[(l-ethoxyethoxy)]propan-1-ol 73 as a colourless oil (2.78 g, 90%) b.p. 49.5-50°C/0.04 mmHg (lit. 68 77-79°C/17 Torr). IR Vmax (film):

3400, 3360, 2900, 1430, 1370, 1320, 1110, 1035, 955 cm·1. 1H n.m.r. 6 (CDC13, 300

MHz) (major isomer): 1.15, d. J 6.2 Hz, (H3)3; 1.19, t, J 7.2 Hz, (H2")3; 1.33, d, J 5.6

Hz, (H2')3; 2.45, dd, J 6.2, 4.6 Hz, 1-OH; 3.38-3.61, m, (Hl)2, H2, H1 l "; 3.68, dq, J9.2,

7.2 Hz, Hbl"; 4.78, q, J 5.6 Hz, HI'. 6 (minor isomer): 1.10, d, J 6.2 Hz, (H3)3; 1.20, t,

J7.2 Hz, (H2")3; 1.32, d,J5.6 Hz, (H2')J; 3.17, dd,J8.7, 3.6 Hz, 1-OH; 3.38-3.61, m,

Hal, H2, (Hl")i; 3.74-3.86, m, Hbl; 4.70, q, J 5.1 Hz, HI'. 13C n.m.r. 6 (CDCl3, 75.6

MHz) (major isomer): 15.2, CH3; 17.5, CH3; 20.3, CH3; 61.0, Cl"; 66.7, Cl; 72.9, C2;

99.4, Cl'; 6 (minor isomer): 15.2, CH3; 17.2, CH3; 20.6, CH3; 60.6, Cl"; 67.1, Cl; 75.7, Chapter 3 Experimental 112

C2; 99.6, Cl'. Mass spectrum: mlz 148 (M\ absent), 133 (M-15, 3%), 117 (7), 103

(22), 87 (5), 73 (100), 59 (45), 45 (82).

Preparation of (2S)-[(1-ethoxy)ethoxy)propionaldehyde 74

a. Oxidation of alcohol 73

i. with pyridinium chlorochromate (PCC)

Preparation of PCC: Pyridine (10.0 mL, 0.125 mol) was added dropwise to a solution of Cr03 (12.5 g, 0.125 mol) in 6 M HCl at 0°C with stirring to give a yellow orange solid. The orange solid was dried for I h in vacuo to give pyridinium chlorochromate m.p. 205°C (lit. 122m.p. 205°C) which was stored in a brown bottle.

Reaction: A solution of alcohol 73 (0.10 g, 0.68 mmol) in anhydrous CH2Cl2

(3.0 mL) was added in one portion to a stirred solution of pyridinium chlorochromate

(0.22 g, 1.01 mmol) in CH2Cli (3 .0 mL) containing a suspension of 4A molecular sieves at r.t. under argon. After 1.5 h, diethyl ether (10 mL) was added and the supernatant decanted from the black gum. The residue was extracted with three more portions (20 mL) of ether whereupon it became a granular solid. The combined supernatant and extracts were passed through a short pad of Florisil, and the solvent removed to give an oil (0.070 g) whose 1H n.m.r. spectrum revealed the presence of 90% starting material

and 10% aldehyde 74.

1· I 2' ii. by Swern oxidation 2·,.,-...01'? Me'2'CHO Oxalyl chloride (2.35 mL, 27.0 mmol) was dissolved in 3 1 74 CH 2Ch (40 mL) at r.t. The solution was stirred and cooled in a Chapter 3 Experimental 113

dry ice/methanol bath at -60°C, then dried DMSO (3.8 mL, 54.0 mmol) in CH2Cli (10 mL) was added dropwise over ea. 10 min. Stirring was continued at -60°C for another

10 min before alcohol 73 (2.00 g, 13.5 mmol) in CH2Cl2 (13.5 mL) was added dropwise over ea. 10 min. The reaction mixture was stirred for 15 min at -60°C, Et3N (9.4 mL,

67 .6 mmol) was added over ea. 15 min and the mixture was stirred for another 30 min at -60°C. The mixture was quenched with water (30 mL) and the cooling bath was removed. Stirring was continued for ea. 10 min and the organic layer was separated.

The aqueous phase was extracted with CH 2Cl2 (3 x 30 mL), and the organic layers were combined, dried over MgSO4, and evaporated to give a pale yellow oil. The oil was column chromatographed using silica gel to give from the ether fraction, a 50:50 diastereomeric mixture of (2S)-[(l-ethoxy)ethoxy]propionaldehyde 74 as a colourless oil (1.58 g, 80%) (lit.68 b.p. 53-54°C/l 7 Torr). IR Vmax (film): 3400, 2980, 2900, 1730,

1440, 1375, 1055, 955 cm·1• 1H n.m.r. 6 (CDCh, 300 MHz): 1.15, 1.17, 2 x t, 17.2 Hz,

(H2")J; 1.25, 1.29, 2 x d, J 6.7 Hz, (H3)J; 1.34, 1.36, 2 x d, J 5.1 Hz, (H2')J; 3.44-3.71, m, (HI "h; 3.92, dq,J 2.6, 6.7 Hz, H2 (isomer a); 4.13, dq, J 2.0, 7.2 Hz, H2 (isomer b);

4.71, 4.80, 2 x q, HI'; 9.57, d, J 3.1 Hz, CHO (isomer a); 9.61, d, J 1.5 Hz, CHO

13 (isomer b). C n.m.r. 6 (CDC13) (75.6 MHz): 15.0, 15.1, CH3; 15.6, 15.9, CH3; 20.15,

20.22, CH3; 61.1, 61.4, Cl"; 75.1, 77.1, C2, 99.5, 100.4, Cl'; 203.1, 203.4, Cl. Mass spectrum: m/z 146 (M\ absent), 117 (M-29, 4%), 101 (14), 73 (60), 57 (14), 45 (100). Chapter 3 Experimental 114

b. Diisobutylaluminium hydride (DIBAL) reduction of ester 72 i. toluene reagent in toluene

A solution of ester 72 (0.100 g, 0.52 mmol) in dry toluene (5.0 mL) was cooled in a bath at -60°C under argon and diisobutylaluminium hydride in toluene (0.25 mL of

2.5 M, 0.676 mmol) was added dropwise. The solution was kept at -60°C for 2 h before a saturated aq. solution of sodium metabisulfite (5.0 mL) was added. The mixture was allowed to warm to r.t. and the layers were separated. The toluene layer was extracted with more sat. sodium metabisulfite solution (3 x 5.0 mL). The combined aqueous extracts were basified with 2 M NaOH solution to pH 8-9 (with cooling) and the solution extracted with ether. The ether extracts were washed with water, dried, and evaporated to give an oil (0.05 g).The oil was flash chromatographed on silica gel using diethyl ether as solvent to yield a 50:50 diastereomeric mixture of aldehydes 74 as a colourless oil (0.027 g, 35%), identical in all respects to that obtained from the Swem oxidation of alcohol 73.

ii. toluene reagent in CH2Cl2

A toluene solution of diisobutylaluminium hydride (8.4 mL of 2.5 M, 12.6

mmol) was added dropwise to a solution of ester 72 (2.00 g, 10.5 mmol) in dry CH2Cl2

(50.0 mL) in a bath at -78°C under argon. Upon complete addition, the mixture was

stirred for a further hour then a solution of KF (5.2 g) in H20 (2.8 mL) was added. The

reaction mixture was allowed to warm slowly to r.t. to give a white solid. The mixture

was filtered through filter aid and the residue washed with CH2Cl2, The filtrate and Chapter 3 Experimental 115

washings were concentrated in vacua at 35°C to afforded an aldehyde 74 as an oil (2.02 g) that from 1H n.m.r. spectroscopic analysis contained aldehyde 74 (0.80 g, 52%)

iii. hexane reagent in CH2Cl2

A hexane solution of diisobutylaluminium hydride (39.5 mL of 1.0 M, 39.5 mmol) was added dropwise to a stirred solution of ester 72 (5.00 g, 26.3 mmol) in dry

CH2Cl2 (25.0 mL) in a bath at -78°C under argon. Upon complete addition, the mixture was stirred for a further 1.5 h then quenched with water (10 mL). The reaction mixture was allowed to warm slowly to r.t., filtered through filter aid and the residue washed with CH2Cl2• The filtrate and washings were extracted with CH2Cl2 (3 x 25 mL) and the extracts were evaporated to dryness in vacua to afford a 1: 1 diastereomeric mixture of aldehydes 74 as an oil (3.71g). The crude oil was chromatographed on silica gel, eluting from ether to afford the aldehyde 74 (3.42 g, 89%), identical in all respects to that obtained from the Swem oxidation of alcohol 73.

Preparation of ethyl (Z- and E-) (4SJ-[(1 '-ethoxy)ethoxy]pent-2-enoate 76a/76b

Preparation of (ethoxycarbonylmethylene)triphenylphosphorane: Ethyl bromoacetate (8.35 g, 50 mmol) was added to a stirred solution of triphenylphosphine

(13.10 g, 50 mmol) in benzene (60 mL) at r.t. The mixture was stirred for 24 hand the resulting precipitate was filtered, washed with diethyl ether, and dried in vacua to give

(ethoxycarbonylmethylene)triphenylphosphonium bromide as a white powder (20.40 g,

95%). The salt was dissolved in water (800 mL) and the aqueous solution was extracted

with diethyl ether (2 x 200 mL). The aqueous layer was then treated with 15% aq. Chapter 3 Experimental 116

NaOH solution to bring the mixture to pH 8-9. The precipitate was collected, washed

with ether, and dried in vacuo to give ( ethoxycarbonylmethylene )triphenylphosphorane

75 as a powder (13.85 g, 84%) ready for further use.

Following the method of Kang et al.,61 a mixture of 2'

aldehyde Etoto 7 4 (0.51 g, 3.5 mmol) and : 5 fl4'Me (ethoxycarbonylmethylene )triphenylphosphorane 75 (2.42 g, 2 1!...co Et 1 2

6.8 mmol) in methanol (10.0 mL) was stirred at 0°C for 2.5 76a h. The solvent was evaporated and the residue was extracted EtO~O

with light petroleum/diethyl ether (85: 15). The extracts were Me~C02 Et 76b filtered through a pad of silica gel and the filtrate evaporated to dryness to give an 82: 18 mixture of (Z)- and (E)-isomers of diastereoisomeric ethyl

(4S)-[(1-ethoxy)ethoxy]pent-2-enoate 76a/76b as a colourless oil (0.74 g, 99%) (lit.67

65°C/8 mmHg). IR Vmax (film): 2980, 2900, 1720, 1645, 1440, 1380, 1190, 1110, 1050,

1 1 820 cm· • H n.m.r. 6 (CDCl3, 300 MHz): 1.13, 1.16, 2 x t, J7.2 Hz, CH3; 1.23-1.28, m.

CH3; 3.36-3.48, m, 3.53-3.63, m, 2 x CH2; 4.14, q, J 7 .2 Hz, H4 (Z-isomer); 4.34, dq, J

5.1, 2.1 Hz, H4 (£-isomer); 4.62, 4.65, 2 x q, J 5.1, H 1' (Z-isomer); 4. 74, q, J 5.1 Hz,

Hl' (£-isomer); 5.69, 5.75, 2 x dd, J 11.8, 1.6 Hz, H2 (Z-isomer); 5.91, 5.96, 2 x dd, J

13.3, 1.0 Hz, H2 (£-isomer); 6.11, 6.22, 2 x dd, J 11.8, 8.2 Hz, H3 (Z-isomer); 6.83,

6.92, 2 x dd, J 15.9, 6.2 Hz, H3 (£-isomer). 13C n.m.r. 6 (CDCl3, 75.6 MHz): 14.1, 14.5,

Cl"; 60.8, 61.5, C4; 68.4, 68.7, Cl'; 98.9, 99.6, C3; 117.9, 119.5, C2; 152.0, 152.9, CO.

Mass spectrum: mlz 216 (M\ absent), 201 (M-15, 0.2%), 171 (2), 157 (0.2), 144, (7),

127 (41 ), 115 (7), 99 (77), 73 ( 100), 46 (86). Chapter 3 Experimental 117

Preparation of (5S)-methyl-2(5H)-furanone 30

5 __,Me Following the method of Kang et al.,67 aq. sulfuric acid (30%, 8 drops) was added to a stirred solution of 76a/76b (l.78 g, 8.2 mmol) in V, 0

MeOH (8.0 mL) at r.t. The reaction mixture was stirred for 1 h, then 30 concentrated, diluted with water and extracted with CH2Cl2 (3 x 20 mL). The extracts were dried over MgS04 and evaporated in vacuo to yield an oil (1.08 g).

Chromatography of the oil on silica gel, eluting with CH2Cli afforded (5S)-methyl-

1.c, 2(5H)-furanone 30 as a colourless oil (0.58 g, 71 %), [a]o +8.6° (c 3.5, CHC13), (Iit.67 b.p. 98-100°C/12 mmHg, [a]o +8.3° (c 3.3, CHCh)). IR Vmax (film): 1745, 1600, 1445,

1365, 1315, 1160, 1100, 1070, 1020, 955, 885, 810 cm-1• 1H n.m.r. 6 (CDC1 3, 300

MHz): 1.44, d, J 7.2 Hz, CH3; 5.11, m, H5; 6.08, dd, J 5.6, 2.1 Hz, H4; 7.44, dd, J 5.6,

13 1.0 Hz, H3. C n.m.r. 6 (CDC1 3, 75.6 MHz): 18.8, CH3; 79.6, C5; 121.2, C4; 157.4, C3;

173.0, CO. Mass spectrum: mlz 99 (M+l, 2%), 98 (M\ 50), 83 (55), 69 (15), 55 (100),

43 (58).

3.2 Preparation of Michael Donors

3.2.1 Attempted preparation of2-(2-bromomethylphenyl)ethanol 79

a. Oxidation of indene 78

i. with ozone in methanol OR1 cqOR1 A solution of indene 78 (11.6 g, 0.1 mol) in methanol (75 2 OR2 mL) was cooled to -30°C (PhBr/C02). Ozone was generated and

passed through the solution ( oxygen gas flow rate 25 L/h, 150 e V) 84a H Me 85a Me H for 6 h. The gas flow was turned off and dimethyl sulfide (10 mL, 86a Me Me Chapter 3 Experimental I I 8

0.136 mol) was added immediately. The solution was then stirred at -I0°C (NaCl/ice) for 1 h, then at 0°C for 1 h, and finally at r.t. for 1 h. The solvent was removed and the residue was dissolved in CH2Cl2 (200 mL) and the solution washed with water. The organic layer was dried over MgSO4 and evaporated to give a pale yellow gum (9.50 g).

N.m.r. spectroscopic analysis revealed that the gum contained a minor amount of diadehyde 80 and a more major mixture of acetals with presumed structure 84a, 85a and

86a.

ii. with ozone in dry CH2C/2

Following the method of Garratt and Vollhard, 79 distilled indene 78 ( 10.0 mL,

9.8 g, 85 mmol) was added to dry dichloromethane (200 mL), the solution cooled to -

65°C, and ozone ( oxygen flow rate 1 L/min, l 50e V) was bubbled through the solution for 6 h. The solution was flushed with Ar (- 10 min), and then acetic acid (20 mL) and zinc (4.0 g) were added and the solution allowed to warm to 0°C with stirring. Four similar portions of zinc in acetic acid were added over the next 2.5 h. The resulting mixture was then filtered, and the filtrate washed with 2M aqu. Na2CO3 solution (150 mL) and water (3 x 100 mL), and then dried over MgSO4• Evaporation of the solvent in vacuo gave a pale yellow oil (7.85 g). 1H n.m.r. spectroscopy showed only the presence of starting material 81.

iii. with KMnO4 (X> .... QH

A solution of KMnO4 (2.72 g, 17.2 mmol) in water (35.0 mL) ''OH

83 was added to a vigorously stirred solution of indene 78 (1.00 g, 8.6 Chapter 3 Experimental 119

mmol) in THF (50.0 mL) over 20 min. The temperature of the mixture increased to

40°C during the addition. After the addition was complete, the mixture was left for another 0.5 h at ambient temperature. The solid that separated was filtered off and the filtrate was concentrated and extracted with diethyl ether to afford a brown oil (0.70 g).

The oil was Kugelrohr distilled to give a white solid (0.26 g) b.p. 150°C/0. l mmHg that was recrystallized from diethyl ether to afford cis-indane-1,2-diol 83 as white needles

(0.089 g, 7%) m.p. 102-103°C (lit.84 m.p. 107-108°C) (Found: C, 72.03; H, 6.52.

C9H10O2 requires: C, 72.07; H, 6.72%). IR Vmax (Nujol): 3320, 3260, 2900, 1450, 1370,

1335, 1175, 1160, 1095, 1045, 995, 735cm·1. 1H n.m.r. 6 (CDC13, 500 MHz): 2.75, brt, J ea. 4 Hz, OH (exchanged with D2O); 2.87, brt, J ea. 6 Hz, OH (exchanged with D2O);

2.93, dd, J 16.3, 3.6 Hz, H8 3; 3.09, dd, J 16.3, 5.8 Hz, Hb3; 4.44, brs, H2 (ddd, J 5.1 Hz, after D2O-exchange); 4.95, dd, J 5.1, 5.0 Hz, HI (d, J 5.1 Hz, after D2O-exchange);

7.20-7.30, m, H4-H6; 7.41, d, J 8.3 Hz, H7. 13C n.m.r. 6 (CDCh, 75.6 MHz): 38.6, C3;

73.5, C2; 76.0, Cl; 125.1, 125.4, 127.2, 128.8, C4-C7; 140.2, 142.0, C3a and C7a. Mass

spectrum: mlz 150 (M\ 25%), 132 (52), 119 (18), 107 (71), 104 (100), 91 (50), 77 (65),

65 (27), 51 (34).

b. Attempted oxidation of eis-indane-1,2-diol 83 with NalO4

i. in 95% ethanol with 5M H2S04

A solution of sodium periodate (NalO4) (0.143 g, 0.67

mmol) in SM sulfuric acid (7.7 mL) at 20°C was added rapidly to a

solution of eis-indane-1,2-diol 83 (0.10 g, 0.67 mmol) in 95% 81 92 ethanol (10 mL) at 40°C. After 10 min the brown solution was 84b H Et 85b Et H 86b Et Et Chapter 3 Experimental 120

cooled to 15°C, diluted with water (10 mL) and extracted with CH2Cl2 (3 x 10 mL). The extracts were dried over MgSO 4 and evaporated to dryness in vacuo to afford a brown oil (0.11 g). N.m.r. spectroscopic analysis showed that the oil contained a major mixture of acetals with presumed structure 84b, 85b and 86b, which could not be purified by column chromatography on silica gel.

A solution of sodium periodate (NalO4) (0.713 g, 3.33 mmol) CCCHO

CHO in SM H2SO4 (50 mL) at 20°C was added rapidly to a solution of cis- 80 indane-1,2-diol 83 (0.50 g, 3.33 mmol) in THF (50 mL) and water (50 mL) at 4CfC. After 5 min, the colourless solution was cooled to 15°C, then water (20 mL) was added to the mixture and the mixture was extracted with CH2Cl2 (3 x 30 mL).

The extracts were washed in tum with 5% aqu. NaiS2O5 solution, 5% aq. NaHCO3 solution and water. The solution was dried over MgSO4 and evaporated in vacuo to afford a brown foam (0.218 g).The 1H n.m.r. spectrum of the crude product revealed the presence of homophthalaldehyde 80, 1H n.m.r. 6 (CDCl3, 300 MHz): 4.15, s,

CH 2CHO; 7.25,m, lxArH; 7.57, ddd,J7.2, 7.2, 1.6Hz,2xArH; 7.84,dd,J7.2,2.1 Hz, lxArH; 9.81, t, J 1.0 Hz, CH2CHO; 10.05, s, CHO, but the substance polymerized after storage in the freezer for 2 days. Chapter 3 Experimental 121

c. via Claisen rearrangement of allyl ether 87

Bromination of 3-methylphenol

Freshly recrystallized ~Br Br~ ~3 NBS (3.46 g, 19 mmol) HoY,cH3 HO~CH3 HOMCH3 Br was rapidly added to a 94 95 96 solution of 3-methylphenol (2.00 g, 18 mmol) in carbon tetrachloride (CC14) (100 mL) followed by benzoyl peroxide (0.0lg, 0.04 mmol). The mixture was refluxed under white light irradiation for 3 h then cooled, filtered under suction and the filtrate and

CC1 4 washings were evaporated in vacuo to afford a colourless oil (4.42 g). 1H n.m.r. spectroscopic analysis revealed the oil to comprise a 1: 1.5: 1 mixture of 2-, 4-, and 6-

1 bromo-3-methylphenol 94, 95 and 96, which was not separated. H n.m.r. o(CDC13,

300 MHz) (2-bromo-3-methylphenol 94): 2.41, s, CH3; 5.53, s, OH; 6.81, d, J 7.7 Hz,

H6; 6.87, d, J 6.7 Hz, H4; 7.11, dd, J 7.7, 7.7 Hz, H5; o (4-bromo-3-methylphenol 95):

2.33, s, CH3; 5.63, s, OH; 6.55, dd, J 8. 7, 3.1 Hz, H6; 6. 73, d, J 3.1 Hz, H2; 7.35, d, J

8.7 Hz, H5; o (6-bromo-3-methylphenol 96): 2.29, s, CH3; 5.51, s, OH; 6.62, dd, J 8.2,

2.1 Hz, H4; 6.85, d, J 1.5 Hz, H2; 7.32, d, J 8.2 Hz, H5. o 13C (CDC13, 75.6 MHz) (2- bromo-3-methylphenol 94): 23.1, CH3; 106.8, C2; 113.2, C4; 122.5, C6; 128.1 CS;

138.6, C3; 152.3, Cl. o (4-bromo-3-methylphenol 95): 21.5, CH3; 106.9, C4; 115.3, C2;

122.8, C6; 130.5, C5; 139.1, C3; 152.5, Cl. o (6-bromo-3-methylphenol 96): 21.0, CH3;

106.8, C6; 116.7, C4; 122.7, C2; 131.5, C5; 139.6, C3; 152.0, Cl. Chapter 3 Experimental 122

Acetylation of 3-methylphenol 3-Methylphenol 88 (10.00 g, 92.4 mmol) was dissolved in 0 acetic anhydride (120 mL) containing pyridine (20 drops). The AcO CHa

97 mixture was stirred at r.t. for 72 h. The mixture was then diluted

with water (150 mL) and the two-phase mixture stirred vigorously for 30 min. The

solution was extracted with EtOAc (3 x 100 mL) and the extracts washed sequentially

with ice-cold 6M HCl (20 mL), brine and sat. NaHCO3 (5 x 100 mL), water, and dried

over MgSO4. The solvent was evaporated under reduced pressure and the residual pale

yellow oil (12.35 g) Kugelrohr distilled to afford m-tolyl acetate 97 as a colourless oil

(11.17 g, 81%) b.p. 80-81°C/0.05 mmHg (lit.123 b.p. 88°C/ll mmHg). IR Vmax(film):

3400 (br), 2980 (br), 1760, 1610, 1585, 1485, 1430, 1365, 1200, 1135, 1015, 1000, 940,

780, 685 cm·1• 1H n.m.r. o(CDC1 3, 300 MHz): 2.30, s, COCH3; 2.38, s, 3-CH3; 6.91, d,

J 7.7 Hz, H6; 6.92, s, H2; 7.16, d, J 7.7 Hz, H4; 7.27, t, J 7.7 Hz, H5. 13C n.m.r. o

(CDC13, 75.6 MHz): 21.0, COCH3; 21.1, 3-CH3; 118.5, C2; 122.1, C4; 126.6, C6;

129.1, C5; 139.5, C3; 150.6, Cl; 169.5, ,COCH3. Mass spectrum: mlz 151 (M+l, 3%),

150 (M\ 17), 137 (3), 129 (7), 111 (5), 108 (100), 97 (7), 77 (23), 69 (22), 55 (15).

g'(;t. Bromination of m-tolyl acetate 97 Freshly recrystallized NBS (7 .82 g, 44 mmol) was rapidly 0 AcO~CH Br added to a solution of m-tolyl acetate 97 (5.97 g, 40 mmol) in CCl4 2 98 (600 mL) containing benzoyl peroxide (0.01 g, 0.04 mmol). The

mixture was refluxed under white light irradiation for 3 h then cooled, filtered and the

filtrate and CC14 washings evaporated to dryness to afford a colourless oil (7.36 g).The Chapter 3 Experimental 123

oil was Kugelrohr distilled to give in tum~ 3 '-bromomethylphenyl acetate 9 8 as a ,,..'I g':>

C9H11O2Br requires: C, 47.19; H, 3.96 %). IR Vmax (film): 1755, 1610, 1585, 1480,

1 1 1440, 1365, 1195, 1140, 1005, 950, 905, 890, 785, 685 cm· • H n.m.r. o (CDC1 3, 300

MHz): 2.30, s, COCH3; 4.47, s, CH2Br; 7.03, d, J 8.2 Hz, H6; 7.14, s, H2; 7.25, d, J 7.7

13 Hz, H4; 7.34, t,J 7.7 Hz, H5. C n.m.r. 8 (CDCh, 75.6 MHz): 21.1, CO~H3; 32.5,

CH2; 121.6, C2; 122.2, C4; 126.4, C6; 129.8, CS; 139.3, C3; 150.8, Cl; 169.2 CO.

Mass spectrum: m/z 230 (M+(81 Br), 14%), 229 (3), 228 (M+(79Br), 13), 186 (40), 149

(32), 107 (100), 89 (13), 78 (77), 63 (20), 51 (46), 43 (79).

Rta.,g~.. 3 '-Dibromomethylphenyl acetate 99 as a colourless oil .0

(0.60 g, 10%) b.p. 123-l 25°C/0.02 mmHg. IR Vmax (film): 1755, AcO # CHBr2

1610, 1585, 1480, 1440, 1365, 1195, 1140, 1005, 950, 905, 890, 99

1 1 785,685 cm· • H n.m.r. 8 (CDC1 3, 300 MHz): 2.31, s, CH3; 6.62, s, CHBr2; 7.08, dd, J

• 13 s: 7.2, 2.0 Hz, H6; 7.35, dd, J 5.1, 2.6 Hz, H4 and H6, 7.39, d, J 2.6 Hz, H2. C n.m.r. u

(CDC13, 75.6 MHz): 21.1, CH3; 39.7, CHBr2; 120.1, CS; 123.1, C4; 123.7, C6; 129.6,

C2; 143.1, C3; 150.4, Cl; 169.0, COCH3• Mass spectrum: m/z 310 (M+(81 Br), 4%), 309

(42), 308 (M+(81 Br, 79Br), 9), 307 (82), 306 (M+(79Br)2, 5), 305 (43), 267 (48), 265

(100), 263 (47), 236 (17), 227 (13), 187 (35), 185 (40), 157 (30), 155 (30), 131 (6), 105

(30), 75 (28), 62 (7), 50 (39). Chapter 3 Experimental 124

Attempted preparation of 3-bromomethylphenol 90

Hydrochloric acid ( 11 mL of 2M) was added to a solution of 3-bromomethylphenyl acetate 98 (1.00 g, 7.25 HOMH,OMe mmol) in MeOH (20 mL) at r.t. The mixture was refluxed for 16 100

h, then concentrated in vacuo, and the residue was diluted with H2O (20 mL) and extracted with diethyl ether (3 x 20 mL). The combined extracts were dried over

Na2SO4 and evaporated to dryness to afford a brown oil (0.48 g), which was Kugelrohr Ue.. distilled to give 3-(methoxymethyl)phenol 100 as a colourless oil (0.31 g, 52%) b.p. l 12-l 13°C/0.05 mmHg. IR Vmax (film): 3320 (br), 2920, 1675, 1590, 1445, 1365, 1265,

1 1 1215, 1185, 1145, 1065, 940, 855, 775, 745, 685 cm· • H n.m.r. o (CDC13, 300 MHz):

3.40, s, CH3; 4.44, s, CH2; 6.27, br s, OH; 6.76, dd, J 2.6, 2.6 Hz, H2; 6.85, dd, J 7.7,

13 s:,: 2.1 Hz, H4 and H6; 7.19, dd, J 7.7, 7.7 Hz, H5. C n.m.r. u (CDC13, 75.6 MHz): 58.0,

CH3; 74.5, CH2; 114.7, C6 or C4; 115.0, C4 or C6; 120.0, C2; 129.7, C5; 139.5, C3;

156.1, Cl. Mass spectrum: m/z 139 (M+l, 10%), 138 (M\ 98), 137 (45), 121 (28), 108

(92), 107 (100), 95 (33), 77 (58), 65 (18), 51 (19), 45 (21).

Preparation of allyl ether 101 Anhydrous K2CO3 (0.28 g, 2.00 mmol) and then allyl (0 O~CH20Me bromide (0.25 g, 2.00 mmol) were added to a vigorously 101 stirred solution of 3-(methoxymethyl)phenol 100 (0.28 g, 2.00 mmol) in acetone (AR grade) (20.0 mL) at r.t. The mixture was refluxed under argon for 24 h with continuous vigorous stirring, and then cooled to r.t. The suspension was filtered, the solid washed with fresh acetone and the filtrate and acetone washings concentrated in vacuo. The Chapter 3 Experimental 125

concentrate was dissolved in light petroleum (20 mL) and the solution washed with

Claisen's alkali (20 mL). The organic layer was washed with H2O (3 x 20 mL), dried over MgSO4, and then evaporated to afford a pale yellow oil (0.28 g). The oil was subjected to flash chromatography on silica gel using EtOAc-light petroleum (1 :4) as eluent to give ally/ 3-(methoxymethyl)phenyl ether 101 as a colourless oil (0.27 g, 75%) b.p.195°C (Found: C, 71.12; H, 7.56. CllH14O2·0.4H2O requires: C, 71.25; H, 8.03%).

IR Vmax (film): 3400 (br), 2880 (br), 1585, 1490, 1255, 1150, 1090, 1020, 990, 920, 770,

690 cm ·1. 1H n.m.r. o(CDCIJ, 300 MHz): 3.39, s, OCH3; 4.44, s, CH2OCH3; 4.55, ddd,

J 5.7, 1.5, 1.5 Hz, OCH2CH=CH2; 5.29, ddt, J 10.8, 1.5, 1.5 Hz, OCH2CH=ClfsynHanti;

5.42, ddt, J 17.4, 1.5, 1.5 Hz, OCH2CH=CHsynl:Ianti; 6.06, ddt, J 17.4, 10.8, 5.6 Hz,

OCH2CH=CH2; 6.85, ddd, J 8.2, 2.6, 2.6 Hz, H6; 6.92-6.93, m, H2 and H4; 7.26, dd, J

8.2, 8.2 Hz, H5. 13C n.m.r. o (CDC13, 75.6 MHz): 58.1, OCH3; 68.8, OCH2CH=CH2;

74.5, CH2OCH3; 113.8, C6; 114.1, C4; 117.6, CH=CH2; 120.1, C2; 129.4, CS; 133.3,

CH=CH2; 139.8, Cl; 158.8, C3. Mass spectrum: m/z 179 (M+l, 10%), 178 (85), 163

(5), 148 (67), 133 (30), 117 (40), 107 (37), 91 (39), 78 (100), 63 (21).

Claisen rearrangement of allyl ether 101

i. in N,N-dimethylani/ine:

A solution of N,N-dimethylaniline (2.00 mL) and allyl ether 101 (0.10 g, 0.56

mmol) was heated in a bath at 190°C for 6 h, then the solution cooled to r.t. The mixture

was diluted with diethyl ether (10 mL) and the solution washed in tum with 3M HCl (3

x 10 mL), 5% Na2CO3 solution (10 mL), water and brine, then dried over MgSO4 and

evaporated to dryness to afford a brown oil (0.14 g). 1H n.m.r. spectroscopic analysis Chapter 3 Experimental 126

showed the oil to contain mostly ally) ether 101 with trace of new products 102, 103 and

104.

ii. neat

The ally) ether 101 (0.12 g, 0.67 ~ HO·n mmol) was sealed in an ampoule and the '(;OMe HO~OMe ampoule heated in a Wood's metal bath at 102 104 198-200°C. Reaction for 4 h afforded a dark brown oil (0.11 g) that comprised three components by 1H n.m.r. spectroscopic analysis. The oil was subjected to column chromatography on silica gel using EtOAc-light petroleum (1 :4) as eluent to afford a

45:55 mixture of phenols 102 and 104 as a pale yellow oil (0.071 g, 59 %). 1H n.m.r. 6

(CDC13, 300 MHz): (Compound 102): 3.39, s, OCH3; 3.50, d, J 6.2 Hz, (H3')i; 4.47, s,

CH2 OCH3; 5.08, m, (HI ')i; 5.49, br s, OH; 6.00, m, H2'; 6.77, d, J 8.2 Hz, H6; 6.93, d,

J7.2 Hz, H4; 7.10, t,J7.7 Hz, H5. 6 (Compound 104) 3.38, s, OCH3; 3.40, d,J6.7 Hz,

(H3')i; 4.40, s, 3-CH2OCH3; 5.10-5.17, m, (Hl')2; 5.94-6.08, m, H2'; 6.81, s, H2; 6.84, dd, J 7.7, 1.6 Hz, H6; 7.08, d, J 7.7 Hz, H5. 13C n.m.r. 6 (Compound 104) (CDCl3, 75.6

MHz): 34.8, C3'; 58.1, OCH3; 74.3, CH2OCH3; 115.1, Cl'; 116.3, C2'; 120.2, C2;

124.9, C4; 127.7, C3; 130.4, C5; 136.4, C6; 138.1, Cl.

Repetition of the reaction for 3 h and 6 h gave similar ratios of the ortho and para rearranged products although reaction for 3 h give some (30%) recovered ally)

ether 101 in addition to the phenols. Chapter 3 Experimental 127

c. via Friedel-Crafts allylation of 3-methoxybenzyl chloride

Allyl chloride ( 1.00 g, 12.8 mmol) was added to a mixture of AlC13 ( 1. 70 g, 12.8 mmol) in chlorobenzene (30 mL) at 0°C under vigorous stirring. The mixture was stirred for 5 min at 0°C then 3-methoxylbenzyl chloride 105 (2.00 g, 12.8 mmol) was added dropwise. The mixture was heated to 40°C for I h and then cooled to r.t., diluted with water (30 mL), extracted with ether (3 x 30 mL), and the extracts were dried over

MgSO4 and evaporated in vacuo to yield a pale yellow oil (3.51 g). The oil was

Kugelrohr distilled to give unchanged chloride 105 as a colourless oil (1.85 g) b.p. 120-

1210C/0.3 mmHg.

d. Reaction of a.,a.'-dibromo-o-xylene 107 with NaCN

.. i. with 1 equiv NaCN

NaCN (0.21 g, 3.79 mmol) and a.,a.'-dibromo-o-xylene 107 CN CCCN (1.00 g, 3.79 mmol) was dissolved in benzene (6.0 mL) and MeOH (3.0 mL) at r.t. with stirring. The mixture was stirred for one day at r.t. and 109 then diluted with H2O (20.0 mL), and the mixture extracted with EtOAc (3 x 20 mL).

The extracts were washed with H2O, dried over MgSO4, and evaporated to dryness in vacuo to afford a 61:39 mixture of a.,a.'-dibromo-o-xylene 107 and a new product,

a.,a.'-dicyano-o-xylene 109 as a green gum (0.63 g).Column chromatography on silica

gel using EtOAc-light petroleum (5:95) as solvent afforded dinitrile 109 as a pale

yellow oil (0.29 g, 48%). 1H n.m.r. 6 (CDCl3, 300 MHz): 3.75, s, (CH2)2; 7.39-7.46, m, :Z>1Cf-/2- 2.'1(.Civ 4xArH. 13C n.m.r. 6 (CDCh, 75.6 MHz): 21.6, ~; 116.5, ~; 129.5, 129.9,

4xArCH; 132.9, 2xArC. Chapter 3 Experimental 128

11. with 4 equiv NaCN

NaCN (0.80 g, 16.0 mmol) and a,a'-dibromo-o-xylene 107 (1.00 g, 3.79 mmol) was dissolved in benzene (6.0 mL) and MeOH (3.0 mL) at r.t. under stirring. The mixture was refluxed for 40 min, cooled to r.t., diluted with H2O (20.0 mL), and extracted with EtOAc (3 x 20.0 mL). The extracts were washed with H2O, dried over

MgSO4, and evaporated to dryness to yield a,a'-dicyano-o-xylene 109 as a pale yellow

1 oil (0.57g, 97%). H n.m.r. o(CDC1 3, 300 MHz): 3.75, s, (CH2)i; 7.39-7.46, m, 4xArH.

3.2.2 Attempted preparation of 2-(2-bromomethylphenyl)acetaldehyde N,N

dimethylhydrazone 77

Preparation of 2-(2-bromomethylphenyl)ethanol 79 i. Reaction of isochroman 110 with 48% HBr

48% Aqueous hydrobromic acid (2.0 mL) was added to isochroman 110 (0.10 g,

0.75 mmol) with vigorous stirring. Stirring was continued at r.t. for 4 d and then the mixture was diluted with H2O (10 mL) and extracted with Et2O (3 x 15 mL). The combined extracts were washed with 10% NaHCO3 solution (2 x 20 mL) and H2O (20 mL), dried over MgSO4 and evaporated to give a pale yellow oil (0.077 g), which by lH n.m.r. spectroscopic analysis was unreacted isochroman 110.

Similar treatment at 50°C for 4 d gave the same result.

ii. Reaction ofisochroman 110 with 55% HI ~OH 55% Hydroiodic acid (2.0 mL) was added to isochroman vv,

110 (0.13 g, 1.0 mmol) at r.t. The mixture was stirred at r.t. for 46 112 Chapter 3 Experimental 129

h, then diluted with water (10 mL) and extracted with EtOAc (3 x 15 mL). The combined organic extracts were washed with 10% aq. Na2S2O5 solution (20 mL) and

H2O (20 mL), then dried over MgSO4 and evaporated to give a pale yellow oil (0.095 g). The oil was subjected to preparative t.l.c., and development with EtOAc-light petroleum (30:70) afforded 2-(2-iodomethylphenyl)ethanol 112 as a pale yellow oil

(0.02 g, 8%). IR Vmax (film): 3340, 1700, 1450, 1150, 1035, 755 cm· 1• 1H n.m.r. 6

(CDC1 3, 300 MHz): 2.49, brs, OH; 2.97, t,J6.7 Hz, (H2)2; 3.96, t, 16.7 Hz, (Hl)2; 4.52, s, CH 2I; 7.18-7.26, m, 3xArH; 7.35, d, J 7.7 Hz, H3'. 13C n.m.r. 6 (CDCh, 75.6 MHz):

4.3, CH2I; 35.5, C2; 62.4, Cl; 127.2, C4'; 128.6, C5'; 130.3, C6'; 130.5, C3'; 136.7,

Cl'; 137.5, C2'. Mass spectrum: m/z 259 (M-3, 9%), 245 (2), 149 (3), 135 (100), 127

(9), 117 (41 ), 104 (61 ), 91 (30), 78 (25), 63 (6), 51 (8).

iii. Reaction ofisochroman 110 with gaseous HBr

Generation of hydrogen bromide:89 Bromine was added dropwise to freshly distilled tetrahydronaphthalene (tetralin) and the resulting gas was passed through a flask containing more tetralin and then into the reaction flask.

Isochroman 110 (2.00 g) was saturated with gaseous HBr in ~Br ~Br the absence of solvent and the mixture kept in a sealed container for 111 17 h. Workup as in part i gave a pale yellow oil (2.86 g).The oil was subjected to column chromatography on silica gel. Elution with EtOAc-light petroleum

( 6:94) gave a fraction, J-bromo-2-(2-bromomethylphenyl)ethane 111 the crystallized as white needles (0.16 g, 4%) m.p. 54-55°C (Found: C, 39.10; H, 3.41. Cg}IrnBr2 requires: Chapter 3 Experimental 130

C, 38.89; H, 3.63%). IR Vmax (Nujol): 2900, 1460, 1390, 1225. 1205, 770 cm· 1• 1H n.m.r. o (CDCl3, 300 MHz): 3.31, t, J 7.7 Hz, (Hl)i; 3.64, t, J 7.7 Hz, (H2)2; 4.57, s,

(Hl")i; 7.22-7.31, m, 3xArH; 7.35, m, H3. 13C n.m.r. o(CDC1 3, 75.6 MHz): 31.2, Cl;

31.7, Cl"; 35.7, C2; 127.7, C4' or CS'; 129.3, CS' or C4'; 130.2, C6'; 130.9, C3';

135.9, Cl'; 137.8, C2'. Mass spectrum: mlz 280 (M+( 81 Br)i, 2%), 278 (M +(81 Br, 79Br),

5), 276 (M+(79Br)i, 2), 200 (4), 199 (39), 198 (5), 197 (37), 117 (100), 104 (20), 91 (25),

78 (9), 63 (12), 51 (12).

Elution with EtOAc-light petroleum (20:80) afforded the ~OH ~Br major component, 2-(bromomethyl)phenylethanol 79 as a pale

79 yellow oil (2.17 g, 68%) b.p. 100-105°C/0.12-0.15 mmHg. IR Vmax

(film): 3280, 2930, 2860, 1435, 1205, 1030, 750 cm·1. 1H n.m.r. O(CDC1 3, 300 MHz):

2.99, s, OH; 3.00, t, J 6.7 Hz, (H2)i; 3.90, t, J 6.7 Hz, CH2OH; 4.58, s, CH2 Br; 7.19-

7.31, m, 3xArH; 7.35, d, J7.7 Hz, H3'. 13C n.m.r. o(CDC1 3, 75.6 MHz): 31.8, C2; 35.3,

CH2Br; 62.8, Cl; 127.1, C4'; 129.1, CS'; 130.4, C6'; 130.7, C3'; 136.2, Cl'; 137.4,

C2'. Mass spectrum: mlz 216 (M+ (81 Br), 2%), 214 (M+(79Br), 2), 134 (79), 117 (26),

104 (100), 91 (21 ), 78 (41 ), 63 (16), 51 (47).

Similar treatment with commercial gaseous HCl for 17 h gave 82% recovery of

starting material but none of the desired chloro alcohol.

Oxidation of 2-(bromomethyl)phenylethanol 79 ~o ~Br A solution of alcohol 79 (0.10 g, 0.47 mmol) in anhydrous n CH2Cl2 (3.0 mL) was added in one portion to a stirred suspension of Chapter 3 Experimental 131

pyridinium chlorochromate (PCC) (0.15 g, 0.70 mmol) and a few 4A molecular sieves

in CH2Cl2 (3.0 ml) at r.t. under argon. After 1.5 h, diethyl ether (10 ml) was added and the supernatant was decanted from the black gum. The insoluble residue was extracted with more ether (3 x 20 ml) until it became granular. The combined supernatant and extracts were passed through a short pad of Florisil, and the solvent was removed to give a brown oil (0.13 g). Preparative t.l.c. using EtOAc-light petroleum (6:94) as developing solvent afforded (2-bromomethylphenyl)acetaldehyde 77 as a highly unstable pale yellow oil (0.04 g, 45%). IR Vmax (film): 2900, 2819, 1720, 1436, 1260,

1200, 1080, 750,650 cm·1• 1H n.m.r. 6 (CDCh, 300 MHz): 3.88, d, J2.l Hz, CH2CHO;

4.47, s, CH2Br; 7.19-7.41, m, 4xArH; 9.80, t, J2.l Hz, CHO. 13C n.m.r. 6 (CDCh, 75.6

MHz): 31.3, Cl"; 47.5, C2; 128.2, C4'; 129.4, C5'; 130.8, C6'; 131.3, Cl'; 131.4, C3';

136.6, C2'; 198.3, Cl. Mass spectrum: mlz 214 (M+2, absent); 212 (M, absent), 185

(6%), 183 (6), 133 (100), 105 (92), 79 (46), 63 (19), 51 (34), 47 (53).

Reaction of aldehyde 77 with N,N-dimethylhydrazine

N,N-Dimethylhydrazine (0.15 ml, 1.93 mmol) was added ~NNMe2

~NHNMe2 to a solution of (2-bromomethylphenyl)acetaldehyde 77 (0.41 g,

114 1.93 mmol) in dry CH2Cl2 (5.00 ml) containing a few 4A

molecular sieves at r.t. under Ar. The mixture was stirred at r.t. for 2 h then filtered, and

the solvent evaporated to yield a yellow foam (0.53 g). The foam was subjected to

chromatography on alumina, eluting with EtOAc-MeOH (20:80) to yield [2-(N',N'

dimethylhydrazinylmethyl)phenyl]acetaldehyde N,N-dimethylhydrazone 114 as yellow

oil (0.21 g, 47%). 1H n.m.r. 6 (CDC13, 300 MHz): 2.71, s, NNMe2; 3.46, s, HNNMe2; Chapter 3 Experimental 132

3.77, d, J 5.6 Hz, CH2CH=N; 5.08, s, NH; 6.56, t, J 5.6 Hz, CH=N; 6.74, s, CH2NH

7.26-7.39, m, 4xArH. 13C n.m.r. 6 (CDCh, 75.6 MHz): 36.7, C2; 43.0 CH=NNMe2;

55.0, HNNMe2; 68.2, CH2NH; 126.5, C4'; 127.0, CS'; 130.8, C6'; 131.1, C3'; 134.8,

Cl; 140.2, CP; 140.3, C2'.

3.2.3 Preparation of 2-(2-(methoxymethyl)phenyl)acetaldehyde N,N

dimethylhydrazone 117

Preparation of 2-(2-(methoxymethyl)phenyl)ethanol l lS

Bromo alcohol 79 (1.95 g) was dissolved in MeOH (30 mL) ~OH ~OMe at r.t. and the solution heated to gentle reflux for 3 h. The methanol 115 was removed under vacuum and the residue dissolved in CH2Cl2 (20 mL), washed with

H2O (3 x 20 mL), dried over MgSO4 and evaporated to dryness to give a colourless oil

(1.48 g). The oil was subjected to column chromatography on silica gel to give from

CH2Cl2 the isochroman 110 (0.39 g), and from EtOAc-CH2Cli (20:80) 2-(2-

(methoxymethy/)pheny/)ethanol 115 as a colourless oil ( 1.17 g, 77%) b.p 105-

1060C/0.03 mmHg (Found: C, 71.98; H, 8.78. C10H14O2 requires: C, 72.26; H, 8.49%).

IR Vmax (film): 3400 (br), 2926, 2875, 1450, 1381, 1190, 1090, 1045, 754 cm· 1• 1H n.m.r. 6 (CDC13, 300 MHz): 2.63, brs, OH; 2.92, t, J 6.7 Hz, (H2)i; 3.41, s, OMe; 3.83,

13 t, J 6.7 Hz, (Hl)i; 4.47, s, CH2OMe; 7.18-7.25, m, 4ArH. C n.m.r. 6 (CDC13, 75.6

MHz): 35.6, C2; 58.1, OMe; 63.5, Cl; 73.2, ,!;;H2OMe; 126.5, C4'; 128.7, CS'; 130.1,

C6'; 130.2, C3'; 136.0, C2'; 138.4, Cl'. Mass spectrum: mlz 166 (M\ 1%), 147 (5),

133 (7), 121 (27), 117 (26), 104 (100), 91 (22), 77 (15). Chapter 3 Experimental 133

Oxidation of 2-(2-(methoxymethyl)phenyl)ethanol l lS

A solution of alcohol 115 (1.00 g, 6.02 mmol) in anhydrous ~o ~OMe CH2Cl2 (6.0 mL) was added in one portion to a vigorously stirred 116 suspension of pyridinium chlorochromate (PCC) (3.25 g, 15.06 mmol) and 4A molecular sieves in CH 2Cli (30 mL) at r.t. under argon. After 2.5 h, diethyl ether (20 mL) was added and the supernatant was decanted from the black gum. The insoluble residue was extracted with more ether (3 x 30 mL) until it became granular.

The combined supernatant and extracts were passed through a short pad of Florisil and the solvent was removed to give a brown oil (0.82 g) that was subjected to column chromatography on silica gel. Elution with 5 :95 EtOAc-CH2 C 12 gave 2-(2-

(methoxymethyl)phenyl)acetaldehyde 116 as a pale yellow oil (0.64 g, 65%) b.p. 68-

700C/0.05 minHg (Found C, 73.26; H, 7.33. C10H 12O2 requires: C, 73.15; H, 7.37%). IR

Vmax (film): 3064, 2927, 2827, 1714, 1456, 1271, 1194, 1084, 752 cm· 1• 1H n.m.r. o

(CDCI3, 300 MHz): 3.33, s, OMe; 3.74, d, J2.l Hz, (H2)i; 4.42, s, CH 2OMe; 7.21-7.36,

13 m, 4xArH; 9.68, t, J 2.1 Hz, CHO. C n.m.r. o (CDC13, 75.6 MHz): 47.9, C2; 57.9,

OMe; 73.4, CH 2OMe; 127.6, C4'; 128.7, CS'; 130.0, C6'; 131.2, C3'; 131.9, Cl';

136.9, C2'; 199.5, Cl. Mass spectrum: m/z 164 (M+, 2%), 150 (3), 131 (30), 121 (74),

104 (100), 91 (74), 84 (31), 77 (65), 65 (45), 51 (55), 45 (40).

Preparation of 2-(-(2-methoxymethyl)phenyl)acetaldehyde N,N- dimethylhydrazone 117

N,N-Dimethylhydrazine (1.18 mL, 15.55 mmol) was added to a solution of aldehyde 116 (0.85 g, 5.18 mmol) in 117 Chapter 3 Experimental 134

absolute ethanol (7.0 mL) at r.t. The mixture was refluxed for 3.5 h, then the solvent was removed in vacuo to afford a yellow oil (0.99 g). Kugelrohr distillation gave 2-((2-

(methoxymethy/)phenyl)acetaldehyde N,N-dimethylhydrazone 117 as a pale yellow oil

(0.93 g, 87%) b.p. 113-115°C/0.16-0.18 mmHg (Found: C, 69.96; H, 8.72; N, 13.26.

C12H 18N2O requires: C, 69.87; H, 8.79; N, 13.58%). IR Vmax (film): 2850, 1595, 1440,

1 1370, 1240, 1180, 1080, 1010, 745 cm·'. H n.m.r. 6 (CDC1 3) (300 MHz): 2.75, s,

NMe2; 3.40, s, OMe; 3.63, d, J 5.6 Hz, (H2)i; 4.51, s, CH2OMe; 6.65, t, J 5.6 Hz,

CH=N; 7.21-7.24, m, 3xArH; 7.35, dd, J 8.2, 2.0 Hz, H6'. 13C n.m.r. 6 (CDC13) (75.6

MHz): 36.6, C2; 43.2, N(CH 3)i; 58.2, OMe; 72.7, .CH2OMe; 126.6, C4'; 128.2, CS';

129.2, C6'; 129.9, C3'; 136.4, C2'; 136.8, C=N; 137.2, Cl'. Mass spectrum: m/z 206

(M\9%), 173(33), 159(17), 144(14), 130(100), 119(37), 115(25), 103(36),91 (37),

77 (15), 58 (14), 43 (41).

3.2.4 Preparation of 2-(2-(phenoxymethyl)phenyl)acetaldehyde 103 and its

hydrazone 121

Preparation of 2-(2-phenoxymethylphenyl)ethanol 119

A mixture of K2CO 3 (6.43 g, 46.5 mmol), phenol (4.37g, ~OH

46.4 mmol) and 2-(2-bromomethylphenyl)ethanol 79 (2.00 g, 9.30 ~OPh

119 mmol) in acetone (20 mL, AR grade) was stirred vigorously at reflux under argon for 18 h. The suspension was cooled, filtered and the filtrate concentrated in vacuo. The concentrate was dissolved in light petroleum (40 mL) and washed with

1Z3 .. Claisen's alkali (20 mL). The organic layer was washed with H2O (3 x 20 mL), dried over MgSO4, and the solvent evaporated to afford a pale brown oil (1.97 g) which was Chapter 3 Experimental 135

subjected to chromatography on silica gel. Elution with EtOAc-light petroleum (20:80) gave the major fraction, which when Kugelrohr distilled gave 2-(2- phenoxymethylphenyl)ethanol 119 as a colourless oil (1.31 g, 62%) b.p. 173-175°C/0.25 mmHg. (Found: C, 76.18; H, 6.82. C 15H 16O2·0.4H2O requires: C, 76.50; H, 7.18%). IR

Vmax (film): 3350 (br), 3040, 2920, 2880, 1590, 1475, 1450, 1380, 1305, 1290, 1225,

1170, 1030, 1005, 875, 750, 690 cm· 1• 1H n.m.r. 6 (CDC1 3, 300 MHz): 1.73, br s, OH;

2.98, t, J 6.7 Hz, (H2)i; 3.91, t, J 6.7 Hz, (Hl)i; 5.08, s, CH2OPh; 7.02-7.12, m, H3",

13 HS" and H4"; 7.25-7.38, m, SxArH; 7.44, d, J 7.2 Hz, H3'. C n.m.r. 6 (CDC13) (75.6

MHz): 35.7, C2; 63.5, Cl; 68.5, ~H2OPh; 114.8, C2" and C6"; 121.2, C4"; 126.8,

C4'; 128.8, CS'; 129.6, C3" and CS"; 129.9, C6'; 130.3, C3'; 135.1, C2'; 137.8, Cl';

158.6, Cl". Mass spectrum: mlz 228 (M\ 3%), 178 (4), 165 (4), 135 (59), 117 (84),

104 (86), 94 ( 100), 91 (82), 77 (85), 65 (90), 51 (42).

Oxidation of 2-(2-phenoxymethylphenyl)ethanol l l9

A solution of alcohol 119 (0.81 g, 3.50 mmol) in anhydrous ~o ~OPh CH 2Cl2 (3.5 mL) was added in one portion to a vigorously stirred 120 suspension of pyridinium chlorochromate (1.53 g, 7.10 mmol) and 4A molecular sieves in CH2Cl2 (9.5 mL) at r.t. under argon. After 2.5 h, diethyl ether (10 mL) was added and the supernatant was decanted from the black gum. The insoluble residue was further extracted with ether (3 x 15 mL) until it became granular. The combined supernatant and extracts were passed through a short pad of Florisil and the solvent removed to give a brown oil (0.66 g). The oil was subjected to column

chromatography on silica gel and eluted with EtOAc-light petroleum (30:70) to afford Chapter 3 Experimental 136

2-(2-(phenoxymethyl)phenyl)acetaldehyde 120 as a pale yellow oil (0.30 g, 38%). 1H n.m.r. 8 (CDCh, 300 MHz): 3.80, d, J 2.1 Hz, CH2CHO; 5.02, CH2OPh; 6.95-7.03,

13 5xArH; 7.26-7.40, 4xArH; 9.77, t, J 2.1 Hz' CHO. C n.m.r. 8 (CDC1 3, 75.6 MHz):

47.8, C2; 68.6, CH2OPh; 114.6, C2" and C6"; 121.2, C4"; 127.8, C4'; 129.1, C5';

129.4, C3" and C5"; 130.1, C6'; 131.3, C3'; 131.8, Cl'; 135.5, C2'; 158.2, Cl"; 199.3,

Cl.

The aldehyde was unstable and was fully characterised as its N,N dimethylhydrazone 121.

Preparation of 2-(-(2-phenoxymethyl)phenyl)acetaldehyde N,N-dimethylhydrazone

121

N,N-Dimethylhydrazine (0.31 mL, 3.99 mmol) was added ~NNMe2 ~OPh to a solution of aldehyde 120 (0.30 g, 1.33 mmol) in absolute

121 ethanol (3.0 mL) at r.t. The solution was heated to reflux for 3.5 h, and then the solvent removed in vacuo and the residual yellow oil (0.35 g) was chromatographed on silica gel, eluting with EtOAc-light petroleum (20:80), to give 2-(

(2-(phenoxymethyl)phenyl)aceta/dehyde N,N-dimethylhydrazone 121 as a pale yellow oil (0.22 g, 63%) b.p. 178-79°C/18 mmHg (Found: C, 76.09; H, 7.51; N, 10.44.

C17H20N2O requires: C, 76.27; H, 7.52; N, 10.33%). IR (film) Vmax: 2940, 2850, 1590,

1470, 1450, 1375, 1290, 1230, 1170, 1025, 1050, 880, 805, 750, 690 cm ·1• 1H n.m.r. o

(CDC1 3, 300 MHz): 2.73, s, NMe2; 3.68, d, J 5.6 Hz, (H2)i; 5.13, s, CH2OPh; 6.69, t, J

5.6 Hz, HI; 6.96-7.02, m, 3 x ArH,; 7.26-7.33, m, 5 x ArH; 7.48, d, J 6.7 Hz, H6'. 13C n.m.r. 8 (CDCh) (75.6 MHz): 36.9, C2; 43.1, NMe2; 68.1, CH2O; 114.8, C2" and C6"; Chapter 3 Experimental 137

120.9, 126.8, 128.5, 129.2, C3'-C6'; 129.5, C3" and C5"; 130.1, C4"; 135.3, C2';

135.9. Cl; 137.2, Cl'; 158.8, Cl". Mass spectrum: m/z 268 (M\ 4%), 222 (4), 195 (5),

175 (83), 159 (17), 130 (100), 115 (22), 103 (27), 77 (10), 65 (20).

3.3 Michael Addition

3.3.1 Reactions of enamine 131 with lactone 30

Preparation of enamine 131

Pyrrolidine (4.15 mL, 49 .8 mmol) was added dropwise to a solution of phenylacetaldehyde 132 (2.00 g, 16.6 mmol) in 131 absolute EtOH (15.0 mL) containing few beads of 4A molecular sieves at r.t. under Ar.

The mixture was refluxed for 3 h, the solution was filtered and the solvent evaporated to afford a brown oil (3.03 g). Kugelrohr distillation afforded (E)-/j-styrylpyrrolidine 131 as a pale yellow oil (2.81 g, 98%) b.p. 138-139°C/0.l mmHg (lit.124 142-146°C/0.9 mmHg). IR Vmax (film): 2925, 2850, 1630, 1595, 1482, 1445, 1370, 1305, 1195, 1175,

1145, 930, 780, 743, 690 cm· 1• 1H n.m.r. 6 (CDC1 3, 300 MHz): 1.93, m, (H3)i and

(H4h; 3.24, m, (H2)2 and (H5)i; 5.12, d, J 13.9 Hz, Ha; 1.96, m, H4'; 7.06, d, J 13.9 Hz,

13 H11 ; 7.16-7.20, m, 5xArH. C n.m.r. 6 (CDC13, 75.6 MHz): 25.2, C3 and C4; 48.9, C2 and C5; 97.4, Ca; 122.8, C11; 123.2, C3'; 128.5, C3' and 5'; 135.8, C2' and C6'; 140.2,

Cl'. Mass spectrum: m/z 174 (M+l, 15%), 173 (M\ 100), 172 (47), 170 (23), 144 (15),

130 (25), 117 (18), 104 (24), 91 ( 12), 77 ( 14), 51 (9), 41 (25).

Repetition of the reaction in toluene solvent in the presence of acetic acid gave the enamine 131 in 85% yield. Chapter 3 Experimental 138

Attempted condensation of enamine 131 and lactone 30

1. with TMSC/

TMSCl (0.17 mL, 1.30 mmol) was added to a solution of lactone 30 (0.106 g,

1.08 mmol) in dry CH2Cl2 (2.0 mL) at r.t. under Ar. After 10 min, the enamine 131

(0.224 g, 1.30 mmol) in CH2Ch (2.0 mL) was added dropwise to the mixture at r.t.. The mixture changed colour from pale yellow to brown during the addition, and was stirred for a further 20 min, then quenched with sat. NaHCO3 and extracted with CH2Cl2 (2 x

10 mL). The extracts were evaporated to dryness to yield a brown oil (0.246 g), which by 1H n.m.r. spectroscopic analysis was a mixture of lactone 30 enamine 131 and a new bn·~1-mn ~- substance that possibly arose tl:H:9agh4aetien ef enamine 131 w.ita D fS:et. '11 )1,tyl' r, er l C,Oc)3 ) ( ~MJ fYll-ll): 1-bb I m, 4H i 2 .12) t / J b. 1Hl/ 4H; 5. ~I :J IS"-~ 1-12 J fl-1 / b-12/cJ. / J IS- 3 l-ll> ifl.

11. with trimethylsilyl trifluoromethanesulfonate (TMSOTJ)

Lactone 30 (0.029 g, 0.30 mmol) was dissolved in CDC13 (0.5 mL) in an n.m.r. tube and enamine 131 (0.062 g, 0.36 mmol) was added. The tube was shaken and a 1H n.m.r. spectrum recorded periodically. There was no change within 24 h at r.t., and so

TMSOTf (0.05 mL) was added to the mixture at ambient temperature. The tube was shaken and the 1H n.m.r. spectrum recorded immediately. The spectrum showed a very complex mixture. Chapter 3 Experimental 139

3.3.2 Reactions of hydrazone derivatives

Preparation of phenylacetaldehyde N,N-dimethylhydrazone 134

N,N-Dimethylhydrazine (3.00 g, 49.9 mmol) was added to ~NNMe, a solution of phenylacetaldehyde 132 (2.00 g, 16.6 mmol) in V absolute ethanol (15 mL) at r.t. The mixture was refluxed for 24 134 h, and the solvent was removed in vacua to afford a yellow oil (3 .11 g). The oil was

Kugelrohr distilled to give phenylacetaldehyde N,N-dimethylhydrazone 134 as a pale yellow oil (2.69 g, 99%) b.p. 65-66°C/O.OlmmHg. IR Vmax (film): 2840, 2000, 1600,

1440, 1240, 1130, 1015, 735,690 cm-1• 1H n.m.r. 6 (CDC13, 300 MHz): 2.77, s, Me2;

3.58, d, J 5.6 Hz, CH2; 6.70, t, J 5.6 Hz, CH=N; 7.20-7.32, m, 5xArH. 13 C n.m.r. 6

(CDC1 3, 75.6 MHz): 39.6, CH2; 43.2, Me2; 126.3, C2' and C6'; 128.5, C4'; 128.8, C3' and C5'; 136-.8, C=N; 138.8, Cl'. Mass spectrum: m/z 163 (M+l, 11%), 162 (M\ 91),

14 7 (7), 117 ( 42), 103 ( 15), 91 ( 100), 77 (21 ), 65 (290, 43 (75).

Reaction of phenylacetaldehyde N,N-dimethylhydrazone 134 with methyl crotonate

Phenylacetaldehyde N,N-dimethylhydrazone 134 (0.337 g, 2.04 mmol) in dry THF (2.0 mL) was added dropwise to a

stirred solution of n-BuLi solution (1.2 mL of 1.9 M in hexanes, 135 2.24 mmol) in THF (5.0 mL) at -78°C under Ar. The mixture

was stirred for 15 min at -78°C whereupon its colour changed to brown. Cul (0.078 g,

0.41 mmol) was added rapidly to the mixture at -78°C and the mixture was stirred for

another 1O min to give a dark brown mixture. Methyl crotonate (0.2 g, 2.04 mmol) in Chapter 3 Experimental 140

THF (2.0 mL) was slowly added to the mixture at -78°C, then the mixture was stirred for another 30 min. The mixture was quenched with water (5.0 mL) and allowed to warm to r.t. The mixture was filtered through filter aid, the filtrate was extracted with

CH2Cl2 (3 x 30 mL), the extracts dried over MgSO4 and the solution evaporated in vacuo to afford a brown oil (0.356 g). The residue was purified by column chromatography on silica gel. Elution with EtOAc/light petroleum ( 1: 1) gave a 65 :35 mixture of two diastereomers of methyl 5-(N,N-dimethyl)hydrazono-3-methyl-4- phenylpentanoate 135 (0.057 g, 11%) (Found: C, 68.79; H, 8.73; N, 10.36. C15H22N2O2 requires: C, 68,67; H, 8.45; N, 10.68%). IR Vmax (film): 2940, 2100, 1725, 1595, 1435,

1 1 1225, 1155, 1000, 695 cm· • H n.m.r. 6 (CDC13, 500 MHz): (major) 0.79, d, J 6.6 Hz,

3-Me; 2.15,dd, J 15.4, 8.8 Hz, H82; 2.62, dd, J 15.4, 4.8, Hz, Hb2; 2.54, m, H3; 2.71, s,

NMe2; 3.21, dd, J9.8, 7.2 Hz, H4; 3.64, s, OMe; 6.64, d, J7.4 Hz, HS; 7.21, m, 3xArH;

7.29, m, 2xArH. 6 (minor) 1.03, d, J 6.6 Hz, 3-Me; 1.95, dd, J 15.4, 9.9 Hz, H82; 2.27, dd, J 15.4, 4.0 Hz, Hb2; 2.54, m, H3; 2. 73, s, NMe2; 3.27, dd, J 9.6, 7.4 Hz, H4; 3.58, s,

OMe; 6.68, d, J 7.4 Hz, HS; 7.21, m, 3xArH; 7.29, m, 2xArH. 13C n.m.r. 6 (CDCh, 125

MHz): (major) 18.2, 3-Me; 33.9, C3; 39.6, C2; 43.0, NMei; 51.4, OMe; 55.5, C4;

126.4, C4'; 128.1, C3' and CS'; 128.5, C2' and C6'; 139.0, CS; 141.5, Cl'; 173.7, C2. 6

(minor) 18.0, 3-Me; 34.0, C3; 39.3, C2; 43.1, NMe2; 51.3, OMe; 54.9, C4; 126.5, C4';

128.1, C3' and CS'; 128.6, C2' and C6'; 139.0, CS; 141.5, Cl'; 173.4, CO. Mass spectrum: mlz 263 (M+ I, 2%), 262 (M+, 13), 232 (10), 189 (17), 177 (23), 161 (68), 144

(12), 131 (13), 117 (47), 105 (39), 91 (100), 84 (20), 77 (41), 69 (41) 59 (32), 44 (55).

The 1H and 13C n.m.r. spectroscopic assignments were supported by a C-H correlation experiment. Chapter 3 Experimental 141

Reaction of phenylacetaldehyde N,N-dimethylhydrazone 134 with lactone 30

N,N-Dimethylphenylhydrazone 134 (0.337 g, 2.04 mmol)

in dry THF (2.0 mL) was added dropwise to a stirred solution of n-BuLi (1.25 mL of 1.8 M in hexanes, 2.24 mmol) in THF (5.0 136 mL) at -78°C under Ar. The mixture was stirred for 15 min at -78°C and Cul (0.077 g,

0.4 mmol) was added rapidly to the brown mixture. The mixture was stirred for another

10 min whereupon its colour changed to dark brown. Furanone 30 (0.20 g, 2.04 mmol) in THF (2.0 mL) was slowly added to the mixture at -78°C, then the mixture was stirred for another 30 min. The mixture was quenched with water (5.0 mL), allowed to warm to r.t. and filtered through filter aid. The filtrate was extracted with CH2Cli (3 x 30 mL), and the extracts dried over MgSO4 and evaporated in vacuo to afford a brown oil (0.341 g). Preparative t.l.c. on silica gel using EtOAc/light petroleum ( 1: 1) as developer gave a fraction (Rr 0.36) comprising an 88: 12 mixture of two diastereomers of ( 4S)-(J-phenyl-

2-(N,N-dimethylhydrazono)ethyl)-(5S)-methy/-y-butyrolactone 13 6 as a brown oil

(0.095 g, 18%) (Found: C, 69.02; H, 7.95; N, 10.88. C15H20N2O2 requires: C, 69.20; H,

7.74; N, 10.76%). IR Vmax (film): 2930, 2850, 1760, 1590, 1445, 1250, 1180, 1025, 935,

1 1 760, 700 cm· • H n.m.r. 6 (CDC1 3, 300 MHz): (major) 1.42, d, J 6.2 Hz, 5-Me; 2.16, dd,

J 18.5, 6.2 Hz, H113; 2.52, dd, J 18.5, 9.2 Hz, 8a3; 2.76, s, NMe2; 2.82, m, H4; 3.52, dd,

J9.8, 5.1 Hz, HI'; 4.56, dq,J6.2, 6.2 Hz, H5; 6.59, d,J5.1 Hz, H2'; 7.18-7.37, m,

5xArH. 6 (minor) 0.91, d, J 6.2 Hz, 5-Me; 2.52, dd, J 18.5, 6.2 Hz, H113; 2. 79, s, NMe2;

2.83, m, H4; 2.96, dd,J 18.5, 10.5 Hz, Ha3; 3.47, dd,J9.3, 3.1 Hz, HI'; 4.32, qd,J6.2,

13 6.2 Hz, H5; 6.64, d, J 3.1 Hz, H2'; 7.18-7.36, m, 5xArH. C n.m.r. 6 (CDC13, 75.6

MHz): (major) 20.88, 5-Me; 32.8, C4; 42.9, NMe2; 44.9, C3; 51.6, Cl'; 80.8, CS; 127.3, Chapter 3 Experimental 142

C4"; 128.4, C3" and C5"; 128.9, C2" and C6"; 135.7, C2'; 139.7, Cl"; 176.4, C2.

8 (minor) 20.85, 5-Me; 34.6, C4; 43.1, NMe2; 45.9, C3; 52.7, Cl'; 80.6, C5; 126.4, C4";

128.3, C3" and C5"; 128.5, C2" and C6"; 135.5, C2'; 140.0, Cl"; 176.4, C2. Mass spectrum: mlz261 (M+l, 3%), 260 (M\ 7), 259 (12), 220 (3), 201 (3), 189 (3), 173 (4),

161 (100), 117 (52), 105 (27), 91 (91), 77 (43), 51 (81).

Reaction of hydrazone 117 with methyl crotonate

N,N-Dimethylhydrazone 117 (1.107 g, 5.37 mmol) in dry THF (5.0 mL) was added dropwise to a stirred solution of 1 OMe n-BuLi (3.1 mL of2.ll Min hexanes, 5.91 mmol) in THF (15 mL) at -78°C under Ar. The mixture was stirred for 15 min at - 137

78°C during which time the colour changed to brown then Cul (0.205 g, 1.08 mmol) was added rapidly. The colour darkened further and the mixture was stirred for another

10 min. Methyl crotonate (1.62 g, 16.12 mmol) in THF (5 mL) was slowly added to the mixture at -78°C and stirring was continued for another 30 min. The mixture was quenched at -78°C with sat. NH4Cl solution (10 mL), allowed to warm to r.t., diluted with H2O (20 mL), filtered through filter aid and then extracted with CH 2Cl2 (3 x 30 mL). The extracts were dried over MgSO4 and the solvent was removed in vacuo to afford a brown oil (1.612 g) that by 1H n.m.r. spectrum analysis contained the desired ester 137, ketone 138 and starting material 117 in the ratio 72:19:9. The mixture was subjected to column chromatography on silica gel. Elution with EtOAc-light petroleum

(20:80) gave the major fraction, methyl 5-(N,N-dimethylhydrazono)-4-(2- methoxymethyl)phenyl-3-methylpentanoate 137 as a pale yellow oil (0.539 g, 33%) Chapter 3 Experimental 143

(Found: C, 66.76; H, 8.40; N, 9.12. C 17H26N2O3 requires: C, 66.64; H, 8.55; N, 9.14%).

IR Vmax (film): 2820, 1725, 1600, 1435, 1360, 1250, 1160, 1090, 1005, 745 cm· 1• 1H n.m.r. o(CDCh, 300 MHz): 0.78, d, J 6.4 Hz, 3-Me; 2.20, dd, J 12.1, 8.7 Hz, Ha2; 2.55-

2.65, m, H3; 2.70, s, NMe2; 2.71, dd, J 12.1, 5.6 Hz, Ht,2; 3.36, s, OMe; 3.55, dd, J 1.2

3.4 Hz, H4; 3.65, s, CO2Me; 4.42, d, J 11.7 Hz, CHaHbOMe; 4.61, d, J 11.7 Hz,

13 CHaHbOMe; 6.51, d, J 1.2 Hz, H5; 7.18-7.35, m, 4:xArH. C n.m.r. o (CDC1 3, 75.6

MHz): 18.2, 3-Me; 33.6, C2; 39.8, C3; 43.0, NMe2; 49.9, C4; 51.4, CO2Me; 58.0, OMe;

72.4, CH2OMe; 126.1, C4'; 127.0, CS'; 128.2, C6'; 129.2, C3'; 136.4, Cl'; 140.3, C2';

139.2, Cl; 174.0, CS. Mass spectrum: m/z 307 (M+l, 4%), 306 (M+, 18), 274 (20), 259

(5), 230 (24), 205 (26), 201 (73), 173 (100), 156 (17), 143 (17), 130 (75), 115 (20), 91

(24), 77 (14), 69 (11), 59 (49).

Further elution with EtOAc-light petroleum (20:80) afforded 1-(N ,N-dimethyl) hydrazono-2-(2-methoxymethyl)phenyl-

3-met hylno nan-5-one 138 as a pale yellow oil (0.211, 12%) 138 (HRMS (ES): Found: m/z 355.2386; C20H32N2O2Na requires: mlz 355.2356). IR Vmax

(film): 2880, 2030, 1700, 1445, 1365, 1345, 1070, 1010, 750 cm·1• 1H n.m.r. 6 (CDC13,

300 MHz): 0. 72, d, J 6.8 Hz, 3-Me; 0.86, m, (H8)i; 0.90, t, J 1.2 Hz, (H9)3; 1.26-1.31,

m, (H7h; 1.50-1.55, m, (H4)2; 2.38, t,J7.5 Hz, (H6)i; 2.66, s, NMe2; 2.72-2.77, m, H3;

3.36, s, OMe; 3.49, dd, J, 10.2, 7.5 Hz, H2; 4.42, d, J 12.1 Hz, CHaHbOMe; 4.61, d, J

13 6 12.0 Hz, CHJibOMe; 6.46, d, J 6.8 Hz, HI; 7.16-7.42, m 4:xArH. C n.m.r. (CDCl3,

75.6 MHz): 13.8, C9; 18.5, 3-Me; 22.3, C8; 25.7, C7; 32.3, C4; 42.9, NMe2; 43.3, C6;

48.4, C3; 50.4, C2; 57.9, OMe; 72.4, .CH2OMe; 126.0, C4'; 127.0, CS'; 128.2, C6'; Chapter 3 Experimental 144

129.1, C3'; 136.3, C2'; 140.2, Cl; 140.3, Cl'; 211.3, CS. Mass spectrum: mlz 332 (M+,

6%), 300 (4), 256 (6), 227 (16), 201 (58), 188 (24), 173 (59), 159 (86), 149 (75), 131

(100), 115 (51), 103 (74), 91 (53), 84 (31), 77 (34), 57 (25).

Reaction of hydrazone 117 and lactone 30

N,N-Dimethylhydrazone 117 (0.404 g, 1.96 mmol) in dry

THF (2.0 mL) was added dropwise to a stirred solution of n-BuLi

(1.20 mL of 1.80 M, 2.16 mmol) in THF (5.0 mL) at -78°C under

Ar. The mixture was stirred for 15 min at -78°C whereupon its 139b colour changed to deep red then Cul (0.090 g, 0.47 mmol) was added rapidly and the mixture was stirred for another 10 min. Lactone 30 (0.19 g, 1.93 mmol) in THF (2.0 mL) was slowly added to the dark brown mixture at -78°C and stirring was continued for another 30 min. The mixture was quenched at -78°C with sat. NH4Cl solution (10.0 mL), allowed to warm to r.t., diluted with H2O (20 mL), filtered through filter aid and then extracted with CH2Cfi (3 x 30 mL). The extracts were dried over MgSO4 and the solvent was removed in vacuo to afford a brown oil (0.546 g) that was subjected to column chromatography on silica gel. Elution with EtOAc-light petroleum (1:1) gave a

12:79:9 mixture of three diastereomers as a pale yellow oil (0.54 g, 90%). The mixture was further separated by preparative t.l.c. on silica gel (EtOAc-light petroleum 1: 1) to afford the major isomer (4S)-[l-(2-methoxymethy/)pheny/-2-(N,N- dimethylhydrazono)ethy/J-(5S)-methy/-y-butyrolactone 139b as a pale yellow oil (0.39 g, 65%) (Found: C, 67.31; H, 8.06; N, 8.91. C11H24N2O3 requires: C, 67.08; H, 7.95; N,

9.20%). IR Vmax (film): 2930-2820 (br), 1760, 1445, 1380, 1180, 1080, 1030, 940, 900, Chapter 3 Experimental 145

1 760, 745 cm·'. H n.m.r. 6 (CDC13, 300 MHz): (major) 1.42, d, J 6.7 Hz, 5-Me; 2.13, dd,

J 18.5, 5.1 Hz, H113; 2.44, dd, J 18.5, 9.2 Hz, Ha3; 2.74, s, NMei; 2.84-2.93, m, H4;

3.38, s, OMe; 3.92, dd, J 7.7, 5.1 Hz, HI'; 4.35, d, J 11.3 Hz, CHaHbOMe; 4.58, d, J

11.3 Hz, CHJ:lbOMe; 4.63, qd, J 6.7, 4.1 Hz, H5; 6.52, d, J 5.1 Hz, H2'; 7.18-7.23, m,

2xArH; 7.28-7.30, m, 2xArH. 13C n.m.r. 6 (CDC13, 75.6 MHz): (major) 21.2, 5-; 32.6,

C3; 43.0, N(CH3)i; 44.6, C4; 46.6, Cl'; 58.1, OMe; 73.1, CH2OMe; 81.5, C5; 126.8,

C4"; 127.7, C5"; 128.9, C6"; 130.3, C3"; 136.20, C2'; 136.25, Cl"; 139.6, C2"; 176.6,

C2. Mass spectrum: m/z 305 (M+l, 10%), 304 (M+, 44), 273 (35), 228 (33), 205 (23),

185 (10), 173 (95), 156 (47), 141 (16), 130 (100), 115 (51), 103 (24), 91 (23), 58 (21),

43 (62).

N.m.r. spectroscopic assignments were supported by HSQC, HMBC and NOESY experiments in CDC13•

Reaction of hydrazone 121 with lactone 30

1. with 1.1 equiv. ofn-BuLi

N,N-Dimethylhydrazone 121 (0.125 g, 0.47mmol) in dry THF (1.0 mL) was added dropwise to a stirred solution of n-BuLi (0.24 mL of 2.1 M, 0.51 mmol) in THF

(2.0 mL) at -78°C under Ar. The mixture was stirred for 15 min at -78°C then Cul

(0.018 g, 0.09 mmol) was added to the mixture and the mixture stirred for another 10

min. Lactone 30 (0.046 g, 0.47 mmol) in THF (1.0 mL) was slowly added to the pale

yellow mixture at -78°C and stirring was continued for another 30 min. The mixture was

quenched at -78°C with sat. NH4Cl solution (7 mL), allowed to warm to r.t., diluted

with H2O (10 mL), filtered through filter aid and then extracted with CH 2Cli (3 x 15 Chapter 3 Experimental 146

mL). The extracts were dried over MgSO4 and the solvent was removed in vacuo to afford a pale yellow oil (0.15 g).Analysis of the oil by 1H n.m.r. spectroscopy revealed the presence of starting material 121 only.

The above reaction was repeated at r.t. for 1 h after the hydrazone 121 was added to the solution of BuLi at -78°C. It gave the same results as above.

11. with 2.2 equiv. ofn-BuLi and quenched with D 20

N,N-Dimethylhydrazone 121 (0.20 g, 0.75mmol) in dry THF (2.0 mL) was added dropwise to a stirred solution of n-BuLi (0. 78 mL of 2.1 M, 1.64 mmol) in THF

(6.0 mL) at -78°C under Ar. The mixture was stirred for 15 min at -78°C, whereupon its colour changed to deep red, then quenched at -78°C with D2O (7.0 mL) The red colour was consumed, and the mixture was allowed to warm to r.t. and extracted with EtOAc

(3 x 15 mL). The extracts were dried over MgSO4 and the solvent was removed in vacuo to afford a yellow oil (0.19 g) that was subjected to column chromatography on silica gel. Elution with EtOAc-light petroleum (20:80) gave from the major fraction, deuterated hydrazone 104 as a pale yellow oil (0.12 g, 63%). Mass spectrum: m/z 269

(10%), 268 (6), 222 (5), 175 (80), 159 (18), 130 (100), 115 (25), 103 (30), 77 (15).

The above reaction was repeated and the Cul and reactant furanone 30 was added to the mixture under the conditions for the Michael reaction. It gave only the hydrazone 121. Chapter 3 Experimental 147

3.3.3. Reactions of arylmethyl bromide 141 with lactone JO/methyl crotonate

Preparation of 2-(methoxymethyl)bromobenzene 143 A solution of 2-bromomethylphenyl bromide 144 (1.54 g) in cx:r OMe MeOH (20.0 mL) was refluxed for 20 h. The MeOH was evaporated

143 in vacuo and the residue was then dissolved in EtOAc (30 mL), washed with H2O (3x30 mL), dried over MgSO4, and evaporated to afford a colourless oil ( 1.22 g). The substance was Kugelrohr distilled to give 2-

(methoxymethyl)bromobenzene 143 as a colourless oil (1.21 g, 99%) b.p. 66°C/0.06 mmHg. IR Vmax (film): 3075, 2995, 2925, 2825, 1595, 1570, 1470, 1445, 1390, 1230,

1200, 1130, 1110, 1050, 1030, 975, 750 cm·1• 1H n.m.r. 6 (CDC13, 300 MHz): 3.47, s,

OMe; 4.53, s, CH2OMe; 7.15, td, J 7.9, 1.5 Hz, H4; 7.32, t, J 7.1 Hz, H5; 7.46, d,J 7.7

13 Hz, H3; 7.54, d, J 8.2 Hz, H3. C n.m.r. 6 (CDC13, 75.6 MHz): 58.6, OMe; 73.9,

~H2OMe; 122.8, Cl; 127.4, C3; 128.9, C5; 129.0, C4; 132.6, C6; 137.6, C2. Mass spectrum: mlz 202 (M+(81 Br), 9%), 200 (M+(79Br), 9), 169 (28), 157 (4), 121 (100), 105

(5), 91 (15), 89 (19), 77 (10), 63 (11), 50 (10).

2-(Methoxymethyl)phenyl bromide 143 (0.40 g, 1.99 mmol) in Me ~OH dry THF (3.0 mL) was added to magnesium turnings (0.05 g, 1.99 ~OMe

mmol) in THF (5.0 mL) containing a crystal of Ji. The mixture was 145

then heated to reflux until the magnesium dissolved whereupon the pale yellow solution

was cooled to -5°C and treated with acetaldehyde (0.167 mL, 2.99 mmol). The mixture

was then warmed to r.t. for 0.5 h, quenched with H2O (10 mL) and extracted with

EtOAc (20 mL). The extracts were washed with 10% aq. H2SO4 (10 mL) and then H2O Chapter 3 Experimental 148

(3 x 10 mL), and evaporated to afford a colourless oil (0.308 g), which was Kugelrohr distilled to give 1-(2-methoxymethyl)phenylethanol 145 as a colourless oil (0.202 g,

61 %) b.p. 98-99°C/0.06 mmHg (Found: C, 72.35; H, 8.74. C10H14O2 requires: C, 72.26;

H, 8.49%). IR Vmax (film): 3425, 3200, 3010, 2960, 2850, 1470, 1390, 1300, 1230,

1210, 1130, 1100, 1020, 950, 910, 780 cm·1• 1H n.m.r. B (CDC13, 300 MHz): 1.51, d, J

6.7 Hz, (H2h; 3.17, brs, OH; 3.37, s, OMe; 4.42, d, J 11.3 Hz, CHaHbOMe; 4.61, d, J

11.3 Hz, CHJibOMe; 5.10, q, J 6.7 Hz, HI; 7.27, m, 2xArH; 7.35, dd, J 7.2, 7.2 Hz,

H4'; 7.53, d, J 7.2 Hz, H6'. 13C n.m.r. B (CDC13, 75.6 MHz): 22.7, C2; 58.0, OMe; 66.2,

Cl; 73.4, CH2OMe; 125.6, C6'; 127.3, C3'; 128.9, C4'; 130.1, CS'; 134.6, C2'; 144.5,

Cl'. Mass spectrum: mlz 166 (M+, absent), 151 (M-15, 7%), 148 (63), 134 (43), 119

( 100), 117 (28), 115 (26), 105 (19), 91 (79), 77 (35), 65 (25), 51 (20), 43 ( 51 ).

Preparation of l-(l-bromoethyl)-2-methoxymethylbenzene 141

48% Aqueous HBr (3.0 mL) was added to a solution of 1-{2- ~Br methoxymethyl)phenylethanol 145 (0.15 g, 0.92 mmol) in CH Cl (3 2 2 ~OMe mL) and containing Bu4NHSO4 (0.058 g). The mixture was stirred 141 vigorously for 3 days then diluted with EtOAc (15 mL) and the organic layer washed with H2O (3 x 15 mL), dried over MgSO4, and evaporated to dryness to yield a brown oil (0.177 g). The substance was Kugelrohr distilled to give J-(l-bromoethy/)-2- methoxymethylbenzene 141 as a colourless oil (0.168 g, 81%) b.p. 72-73°C/0.06 mmHg

(Found: C, 52.11; H, 6.11. C10H13BrO requires: C, 52.42; H, 5.72%). IR Vmax (film):

2920, 2875, 2810, 1445, 1375, 1215, 1190, 110, 1090, 1030, 965, 940, 905, 755, 730

cm·1• 1H n.m.r. B (CDCh, 300 MHz): 2.08. d, J 6.7 Hz, CHBrCH3; 3.41, s, OMe; 4.42, Chapter 3 Experimental 149

d, J 11.8 Hz, CHaHbOMe; 4.76, d, J 11.8 Hz, CHaHbOMe; 5.62, q, J 6.1 Hz, CHBr;

7.25-7.31, m, 2xArH; 7.38, td, J 10.2, 2.6 Hz, HS; 7.67, d J 1.1 Hz, H6. 13 C n.m.r. 6

(CDC1 3, 75.6 MHz): 26.4, CHBrCH3; 44.9, CHBr; 58.2, OMe; 72.5, CH 2OMe; 126.9,

C6; 128.1, C3; 128.8, C4; 129.6, CS; 134.5, C2; 141.9, Cl. Mass spectrum: mlz 230

(M+(81 Br), 0.2%), 228 (M+(79Br), 0.2), 198 (2), 196 (2), 149 (28), 133 (13), 117 (100),

103 (12), 91 (26), 77 (15), 63 (8).

Attempted reactions of 1-(1-bromoethyl)-2-methoxymethylbenzene 141 with lactone 30/methyl crotonate i. with n-BuLi

The bromide 141 (0.089 g, 0.39 mmol) in dry THF (2.0 mL) was added to a solution of n-BuLi (0.27 mL of 1.55 M, Meo OMe 0.42 mmol) in THF (6.0 mL) at -78°C under Ar. The mixture 146 was stirred at this temperature for 30 min. Lactone 30 (0.038 g, 0.39 mmol) was added and the mixture stirred for 3 h at -78°C then quenched with H2O (5.0 mL). The mixture was extracted with EtOAc (3 x 10 mL) and the extracts dried over MgSO4 and evaporated to give a brown oil (0.064 g). The oil was subjected to preparative t.l.c. using EtOAc-light petroleum (10:90) to afford, in order, a 65:35 mixture of meso and rac-2,3-bis(2-methoxymethylpheny/)butane 146 as a pale yellow gum (0.019 g, 16%)

(HRMS (ES): Found: m/z 321.1823; C20H26O2Na requires: mlz 321.1825). 1H n.m.r. 6

(CDCh, 300 MHz) (isomer 1): 0.97, d, J 6.2 Hz, CH 3; 3.26, s, OMe; 3.36-3.37, m, H2

and H3; 4.15, d, J 11.8 Hz, CHaHbOMe; 4.38, d, J 11.8 Hz, CHJfbOMe; 6.99, dd, J 1.2,

1.0 Hz, ArH; 7.02-7.22, m, ArH; 7.27-7.40, m, 2xArH. 6 (isomer 2): 1.36, d, J 6.2 Hz, Chapter 3 Experimental 150

CH3; 3.42, s, OMe; 4.52, d, J 11.8 Hz, CHaHbOMe; 4.59, d, J 11.8 Hz, CHaHbOMe;

7.02-7.22, m, 2xArH; 7.27-7.40, m, 2xArH. 13C n.m.r. B (CDC13, 75.6 MHz): 21.1, 21.3,

CH3; 40.1, 40.5, CHCH3; 57.7, 57.9, OMe; 72.8, 73.2, GI2OMe; 125.3, 125.6, C6';

126.3, 126.8, C3'; 127.8, 128.5, C4'; 129.0, 129.5, C5'; 134.6, 135.6, C2'; 146.1, 146.4,

Cl'. Mass spectrum: mlz 298 (M+, absent), 267 (5%), 235 (26), 221 (13), 193 (7), 178

(7), 149 (14), 129 (20), 117 (100), 105 (22), 91 (37), 84 (14).

2-(2-methoxymethylphenyl)hexane 147 as a colourless oil Me ~ (0.013 g, 17%): 0.86, t,J7.7 Hz, (H6)3; 1.20, d,J6.8 Hz, (Hl)3; ~OMe

1.26-1.30, m, (H4)2 and (H5)2; 1.56-1.61, m, (H3)i; 3.01, qt, J

147 7.7, 6.8 Hz, H2; 3.37, s, OMe; 4.45, d, J 11.3 Hz, CHaHbOMe;

4.52, d, J 11.3 Hz, CHJibOMe; 7.13-7.18, m, ArH; 7.27-7.31, m, 3xArH. 13C n.m.r. B

(CDC13, 75.6 MHz): 14.1, C6; 22.4, Cl; 22.9, CS; 30.1, C4; 33.9, C3; 37.9, C2; 57.9,

OMe; 72.9, .CH2OMe; 125.4, C6'; 125.8, C3'; 128.2, C4'; 129.2, CS'; 135.1, C2',

147.0, Cl'. Mass spectrum: m/z 207 (M+l, trace), 175 (30%), 149 (18), 131 (20), 117

(100), 105 (36), 91 (27). The desired product 142 could not be detected by 1H n.m.r.

spectroscopic analysis.

The desired product 142 was also not observed when the reactants were added in

the reverse order or the reaction performed at -100°C.

ii. promoted by Mg

Bromide 141 (0.500 g, 2.18 mmol) in dry THF (3.0 mL) was slowly added to

magnesium turnings (0.053 g, 2.18 mmol) in THF (30.0 mL) containing a crystal ofl2• Chapter 3 Experimental 151

The mixture was heated to reflux until the magnesium dissolved, whereupon it took on a pale yellow colour. Methyl crotonate (0.33 g, 3.28 mmol) was added to the Grignard reagent solution at 0°C and the mixture was warmed to r.t. with stirring over 30 min.

The mixture was diluted with H2O (10 mL) and THF (10 mL) and the organic layer separated. The aqueous layer was further extracted with Et2O (2 x 20 mL), and the combined extracts dried over MgSO4• Evaporation of the solvent gave a pale yellow oil

(0.315 g). 1H n.m.r. spectroscopic analysis indicated the presence of dimer 146 but the absence of the desired product 142.

iii. promoted by indium powder

Water (2 drops) was added to a mixture of bromide 141 (0.071 g, 0.31 mmol), methyl crotonate (0.050 g, 0.50 mmol) and indium powder (0.050 g, 0.44 mmol). The mixture was stirred for 18 h at r.t., diluted with EtOAc, filtered and the filtrate and washings then extracted with EtOAc (3 x 5 mL). The extracts were washed with H2O (3

x 10 mL), dried over Na2 SO 4 and evaporated to give 1-{2-

methoxymethyl)phenylethanol 145 as a colourless oil (0.042 g, 82%). The compound

was identified by comparison of its 1H n.m.r. spectroscopic data with those of an

authentic sample 145.

iv. promoted by n-Bu3SnH

n-Tributyltin hydride (0.14 mL, 0.99 mmol) was added to a solution of bromide

141 (0.114 g, 0.50 mmol), methyl crotonate (0.057 g, 0.57 mmol) and AIBN (0.013 g)

in dry benzene (20 mL) under Ar. The solution was heated to reflux for 4 hand then Chapter 3 Experimental 152

cooled to r.t., quenched with H2O (10 mL) and extracted with EtOAc (2 x 20 mL). The extracts were dried over Na2SO4 and, evaporated to dryness to give a colourless oil

(0.531 g). The oil was subjected to preparative t.l.c. using EtOAc-light petroleum

( 10:90) as developing solvent to afford a major fraction comprising 1-(2- methoxymethyl)phenylethanol 145 as a colourless oil (0.075 g, 83%). The compound was identified by comparison of its 1H n.m.r. spectroscopic data with those of an authentic sample 145.

3.4. Preparation of advanced intermediates

3.4.1. Preparation of aldehyde 155 through solvolysis of lactone 139 i. with EtOH in the presence of30% aq. H2SO4

Lactone 139 (0.207 g) was dissolved at r.t. in

EtOH (2.0 mL) containing 30% aq. H2SO4 (8 drops) and the mixture was warmed to gentle reflux for 5 h. The ethanol was removed in vacuo and the residue diluted 150 with EtOAc (10.0 mL), washed with H2O (2 x 10 mL), dried (NaiSO4) and evaporated to dryness to give a pale yellow oil (0.20 g). The oil was subjected to preparative t.l.c. using 10:90 EtOAc-light petroleum as developing solvent to afford the major fraction, ethyl {5 (R)-ethoxy-(2S)-methyl-(4R)-(2-methoxymethyl)phenyl]tetrahydrofuran-(3S) acetate 150 as a colourless oil (0.072 g, 81%) (Found: C, 67.78; H, 8.53. C19H2sOs requires: C, 67.83; H, 8.39%). IR Vmax (film): 2900 (br), 1725, 1440, 1368, 1155, 1080,

1050, 1010, 970, 750 cm·1• 1H n.m.r. ~ (CDCh, 500 MHz): 1.12, t, J7.l Hz, 5'

OCH2CH3; 1.16, t, J 7.1 Hz, CO2CH2CH3; 1.27, d, J 6.7 Hz, 2'-CH3; 2.42, dd, J9.4, 6.1 Chapter 3 Experimental 153

Hz, H82; 2.45, dd, J 9 .4, 6.8 Hz, Hb2; 2.87, dddd, J 9 .6, 7.1, 6.8, 6.1 Hz, H3 '; 3.31, dq, J

9.6, 7.1 Hz, OCHaHbCH3; 3.38, s, OMe; 3.42, dd, J 9.6, 4.2 Hz, H3; 3.73, dq, J9.6, 7.1

4.65, d, J 11.6 Hz, CHaHbOCH3; 4.62, dq, J 7.0, 7.0 Hz, H2'; 4.97, d, J 4.2 Hz, H5';

7.19-7.35, m, 4xArH. 13 C n.m.r. 6 (CDCh, 75.7 MHz): 14.0, COiCH~H3; 15.2,

OCH~H3; 17.5, 2'-CH3; 33.8, C2; 46.5, C3'; 51.3, C4'; 57.8, OCH3; 60.2,

126.7, C3"; 128.4, C4"; 129.3, C5"; 136.8, Cl"; 139.1, C2"; 172.3, Cl. Mass spectrum: m/z 335 (M-1, trace), 262 (M-74, 6%), 230 (18), 213 (14), 201 (6), 185 (18), 156 (50),

142 (100), 129 (40), 115 (20), 103 (8), 91 (11), 45 (18).

N.m.r. spectroscopic assignments were supported by HSQC and HMBC experiments in

ii. with THFIH2O in the presence of30% aq. HiSO4

Lactone 139b (0.263 g) was dissolved at r.t. in THF (2.0 mL), and H2O (2.0 mL) and 30% aq. H2SO4 (22 drops) were added. The mixture was stirred at r.t. for 3 d and diluted with

EtOAc (10.0 mL), washed with H2O (3 x 10 mL), dried (Na2SO4) 148 and evaporated to dryness to give a pale yellow oil (0.192 g) that was subjected to preparative t.1.c. using EtOAc-light petroleum (50:50) as developing solvent to afford the major fraction as a 63:37 mixture of diastereomeric45-[1-formy/-J-(2- t5) methoxymethy/)phenyl]methy/-5-methy/-y-butyrolactone 148 as a pale yellow oil (0.121 g, 52%) (Found: C, 68.35; H, 6.83. C1sH1sO4 requires: C, 68.69; H, 6.92 %). IR Vmax Chapter 3 Experimental 154

(film):3340,2880,2800, 1750, 1720, 1440, 1410, 1375, 1260, 1190, 1175, 1070, 1030,

935, 750 cm- 1• 1H n.m.r. 8 (CDCh, 300 MHz): (major isomer) 0.96, d, J 6.2 Hz, 5-Me;

2.37, dd, J 17.9, 7.7 Hz, H113; 2.73-2.82, m, H4; 3.00, dd, J 17.9, 8.7 Hz, H0 3; 3.35, s,

OMe; 4.07, d, J 10.7 Hz, CHaHbOMe; 4.28, dq, J 6.2, 6.2 Hz, H5; 4.43, d, J 10.7 Hz,

CHaHbOMe; 7.06, dd, J 8.2, 2.1 Hz, HI'; 7.27-7.37, m, 4xArH; 9.53, brs, CHO.

8 (minor isomer): 1.45, d, J 6. 1 Hz, 5-Me; 2.02, dd, J 17 .9, 9.2 Hz, H113; 2.45, dd, J

17.9, 9.2 Hz, H0 3; 2.92-2.99, m, H4; 3.33, s, OMe; 4.13, d, J 10.7 Hz, CHaHbOMe;

4.41, d J 10.7 Hz, CHJfbOMe, 4.43, dq, J 6.1, 7.2 Hz, H5; 7.00, dd,J 7.2, 2.0 Hz, HI';

7.27-7.37, m, ArH; 9.58, brs, CHO. 13C n.m.r. 8 (CDC13, 75.6 MHz): 8 (major isomer)

20.6, 5-Me; 34.4, C3; 42.7, C4; 57.1, Cl'; 57.9, CH2OCH3; 73.4, CH2OCH3; 79.6, CS;

128.3, C4"; 129.0, CS"; 129.2, C6"; 131.2, C3"; 132.7, C2" 137.3, Cl"; 175.6, C2;

197.8, CHO. 8 (minor isomer) 20.6, 5-Me; 31.4, C3; 41.4, C4; 56.7, Cl'; 57.9,

CH2OCH3; 73.3, CH2OCH3; 81.4, C5; 128.1, C4"; 128.7, CS"; 129.3, C6"; 131.1, C3";

132.7, C2"; 137.5, Cl"; 175.7, C2; 198.5, CHO. Mass spectrum: m/z 261 (M-1, trace),

244 (0.6%), 233 (4), 202 (25), 159 (47), 145 ( 22), 130 (100), 115 (48), 103 (13), 91

(25), 77 (13), 45 (35).

Reaction of aldehyde 148 with ethylene glycol

Ethylene glycol (2.0 mL, 35.9 mmol) was added to aldehyde 148 (0.060 g, 0.229 mmol) in benzene (4.0 mL) and

PPTS (0.069 g) at r.t. under Ar. The mixture was refluxed for 5 h OMe and then cooled to r.t., diluted with EtOAc (10 mL) and washed 149 with water (3 x 15 mL). The organic solution was dried over MgSO4 and evaporated to Chapter 3 Experimental 155

afford a pale yellow gum (0.069 g). The gum was subjected to preparative t.l.c. using

EtOAc-light petroleum (50:50) as developing solvent to give a 54:46 mixture d iastereom eri c of ( 4S )-{]-(dioxo/an-2-yl)-1-(2-methoxymethy/) phenylmethyl]-(5S) methyl-y-butyro/aetone 149 as a pale yellow oil (0.058 g, 83%) (Found: C, 66.30; H,

7.26. C17H22Os requires: C, 66.65; H, 7.24%). IR Vmax (film): 2920, 2865, 1710, 1440,

1370, 1260, 1180, 1170, 1075, 1020, 970, 930, 900, 790, 755, 740. 650 cm·1• 1H n.m.r.

B (CDC13, 500 MHz) (isomer a): 0.84, d, J 6.2 Hz, 5-Me; 2.71, dddd, J 9.8, 9.1, 7.6, 5.7

Hz, H4; 2.75, dd, J 17.4, 5.7 Hz, H 113; 2.78, dd, J 17.4, ea 5 Hz, Ha3; 3.36, s OMe; 3.38, dd, J ea 10, 3.6 Hz, HI'; 3.70-3.80, m, (H4")i and (H5")2; 4.27, d, J 11.3 Hz,

CHaHbOMe; 4.33, dq, J 6.2, 6.2 Hz, H5; 4.66, d, J 11.3 Hz, CHaHbOMe; 5.00, d, J 3.6

Hz, H2"; 7.15-7.35, m, 3xArH; 7.40, m, lxArH. B (isomer b): 1.49, d, J 6.2 Hz, 5-Me;

2.20, dd, J 18.5, 7.7 Hz, H11 3; 2.39, dd, J 18.5, 9.2 Hz, Ha3; 2.78, dddd, J 9.8, 8.9, 8.9,

5.7 Hz, H4; 3.34, s OMe; 3.48, dd, J9.8, 4.0 Hz, HI'; 3.70-3.80, m, (H4")i and (HS")i;

4.23, d, J 11.3 Hz, CHaHbOMe; 4.64, d, J 11.3 Hz, CHaHbOMe; 4. 76, dq, J 6.2, 6.2 Hz,

H5; 5.05, d, J 4.0 Hz, H2"; 7.15-7.35, m, 3xArH; 7.40, m, lxArH. 13C n.m.r. B (CDCl3,

125.7 MHz) (isomer a): 20.8, 5-Me; 34.9, C3; 44.5, C4; 45.8, Cl'; 57.9, OMe; 64.5,

65.4, C4" and CS"; 73.6, ,CH2OCH3; 81.5, CS; 105.8, C2"'; 127.1, C4"'; 128.4, CS"';

128.7, C6"'; 130.4, C3"'; 136.7, Cl"'; 138.1, C2"'; 176.4, C2. 6 (isomer b): 21.3, 5-Me;

33.8, C3; 43.5, C4; 45.8, Cl'; 57.8, OMe; 64.7, 65.2, C4" and CS"; 73.5, ~H2OCH3;

80.9, CS; 105.5, C2"; 127.0, C4'"; 128.5, C5"'; 128.8, C6"'; 130.3, C3"'; 137.0, Cl"';

137.9, C2"'; 176.2, C2. Mass spectrum: mlz 306 (M+, 1%), 274 (6), 201 (3), 185 (2),

175 (6), 163 (11), 160 (10), 149 (8), 145 (14), 142 (17), 141 (24), 73 (100), 65 (39), 55

( 46), 51 (32), 45 (96). Chapter 3 Experimental 156

N.m.r. spectroscopic assignments were supported by HSQC, HMBC and NOESY experiments in CDCI3.

3.4.2. Preparation of sulfide 153

Reduction of aldehyde 148

NaBH4 (0.048 g, 1.28 mmol) was added to a stirred solution Me .::- of aldehyde 148 (0.150 g, 0.53 mmol) in absolute ethanol (10 mL) at r.t. Stirring was continued at r.t. for 1.5 h then the mixture was OMe quenched with water (10 mL). The product was extracted into 151

EtOAc (2xlOmL) and dried over MgSO4• Solvent was evaporated to afford a brown oil

(0.136 g). The oil was subjected to preparative t.l.c. using EtOAc-light petroleum

(70:30) to give ( 4S)-[l-hydroxymethyl-l-(2-methoxymethy/}phenylmethyl]-(5S)-methyl y-butyrolactone 151 as a pale yellow oil (0.112 g, 74%) (HRMS (ES): Found: m/z

287.1256. C15H20O4Na requires: m/z 287.1254). IR Vmax {film): 3450-3360 {br), 2900,

1750, 1370, 1180, 1060, 1045, 930, 750 cm·1• 1H n.m.r. o(CDC1 3, 300 MHz): 1.51, d, J

6.0 Hz, 5-Me; 2.09, dd, J 18.1, 7.1 Hz, H~; 2.41, dd, J 18.1, 8.6 Hz, Ha3; 2.55, m, H4;

3.21, ddd, J 12.4, 8.3, 4.5 Hz, HI'; 3.38, s, OCH3; 3.70, dd, J 10.5, 8.3 Hz, CHaHbOH;

3.87, dd, J 10.5, 4.5 Hz, CHJibOH; 4.41, d, J 10.9 Hz, CHaHbOMe; 4.46, d, J 10.9 Hz,

CHJfbOMe; 4.64, dq, J 6.0, 6.0 Hz, H5; 7.20-7.32, m, 3xArH; 7.36, m, ArH. 13 C n.m.r. o (CDCh, 75.6 MHz): 21.7, CH3; 34.0, C3; 44.5, C4; 45.9, Cl'; 58.0, CH2O~H3; 65.5,

CH2OH; 73.4, CH2OCH3; 80.9, C5; 126.6, C4"; 127.0, C5"; 129.4, C6"; 130.8, C3";

136.3, Cl"; 140.1, C2"; 175.9, C2. Mass spectrum: mlz 264 (M+, absent), 214 (trace), Chapter 3 Experimental 157

202 (27%), 159 (39), 142 (41), 130 (97), 129 (100), 121 (51), 105 (63), 91 (28), 77 (19),

45 (70).

Preparation of tosylate 152

Freshly recrystallized p-toluenesulonyl chloride

(0.041 g, 0.21 mmol) was added to a solution of alcohol

151 (0.028 g, 0.106 mmol) in redistilled pyridine (2.0 mL) at 0°C under stirring. The mixture was stirred at 0°C for 2 hand then placed in a refrigerator 3 d (4°C). The mixture 152 was diluted with 5% aqueous HCl and extracted with CH2Cl2 (2xl5 mL). The combined extracts were washed with water (20 mL), dried over MgSO4 and concentrated in vacuo.

The residue was purified by preparative t.l.c. using EtOAc-light petroleum (50:50) as developing solvent to give (4S)-[ 1-(2-methoxymethyl)phenyl- 2-P- toluenesulfonyloxy·~ ethyl]-(5S)-methy/-y-butyrolactone 152 as a pale yellow oil

(0.035 g, 80%) (HRMS (ES): Found: m/z 441.1343. C22H26O6SNa requires: m/z

441.1342). IR Vmax {film): 2900, 1755, 1595, 1445, 1350, 1170, 1080, 1035, 960,940,

810, 765, 665 cm·1• 1H n.m.r. 6 (CDCh, 300 MHz): 1.46, d, J 6.4 Hz, 5-Me; 2.11, dd, J

18.1, 6.8 Hz, H 113; 2.38, dd, J 18.1, 9.1 Hz, Ha3; 2.41, s, PhCH3; 2.61, dddd, J 10.5, 9.1,

6.8, 5.4 Hz, H4; 3.25, s, CH2OCH3; 3.43, ddd, J 10.5, 7.2, 5.6 Hz, HI'; 4.10, dd, J 9.4,

7.2 Hz, CHaHbOTs; 4.15, dd, J 9.4, 5.6 Hz, CHaHbOTs; 4.26, d, J 11.3 Hz,

CHaHbOCH3; 4.41, d, J 11.3 Hz, CHJibOCH3; 4.59, dq, J 5.4, 6.4 Hz, H5; 7.07,ddd, J

3.0, I.I, 1.1 Hz, lxArH; 7.21-7.26, m, 5xArH; 7.59, dd, J 6.4, 1.5 Hz, 2xArH. 13C n.m.r. 6 (CDC13, 75.6 MHz): 21.4, PhCH3; 21.6, 5-CH3; 33.4, C3; 41.9, C4; 44.1, Cl'; Chapter 3 Experimental 158

57.7, OCH3; 71.7, CH2OCH3; 73.3, gI2OTs; 80.3, CS; 127.1, C4"; 127.3, C5"; 127.7,

C3"' and CS"'; 129.0, C6'; 129.8, C2"' and C6"'; 130.8, C3"; 132.2, C4'"; 136.1, Cl";

138.1, C2"; 145.0, Cl"'; 175.5, C2. Mass spectrum: mlz 419 (M+l, trace), 214 (8%),

186 (10), 155 (21), 146 (49), 128 (48), 116 (43), 105 (68), 91 (100), 77 (16), 65 (32), 45

(51).

Preparation of sulfide 153

Potassium tert-butoxide (0.016 g, 0.143 mmol) was added to a stirred solution of thiophenol (0.015 mL, 0.143 mmol) in dry

DMSO (1.0 mL) under Ar. A solution of the tosylate 152 (0.030 153 g, 0.072 mmol) in DMSO (1.0 mL) was then added and stirring continued at r.t. for lh.

The mixture was diluted with EtOAc (10 mL) and the solution washed with water (2 x

10 mL), dried over MgSO4 and concentrated in vacuo. The residue was subjected to preparative t.l.c. with EtOAc:Iight petroleum (30:70) as developing solvent to afford

(4S)-[ 1-phenylthio ==::, J-(2-methoxy'Jthyl)phenylmethyl]-(5S)-methyl-y butyrolactone 153 (0.021 g, 82%) as a pale brown gum (Found: C, 70.64; H, 7.07.

C21 H24O3S requires: C, 70.76; H, 6.79%). IR Vmax (film): 2880, 1760, 1440, 1350, 1180,

1070, 940,740,690 cm·1. 1H n.m.r. 6 (CDC13, 300 MHz): 1.49, d, J 6.3 Hz, 5-Me; 2.26, dd, J 18.2, 7.6 Hz, H133; 2.36, dd,J 18.2, 8.6 Hz, Ha3; 2.59, dddd,J 12.5, 8.6, 7.6, 5.3

Hz, H4; 3.20, ddd, J 12.5, 5.0, 4.6 Hz, Hl '; 3.24, s, OCH3; 3.28, dd, J 18.2, 5.0 Hz,

CHaHbSPh; 3.38, dd, J 18.2, 4.6 Hz, CHaHbSPh; 4.15, d, J 11.2 Hz, CHaHbOMe; 4.47,

13 s: d, J 11.2 Hz, CHJ!bOMe; 4.56, dq, J 5.3, 6.3 Hz, HS; 7.14-7.34, m, 9xArH. C n.m.r. u

(CDCh, 75.6 MHz): 21.8, S-CH3; 33.2, C3; 39.6, CH2SPh; 41.8, C4; 47.4, Cl'; 57.9,

OCH3; 73.4, .CH2OCH3; 80.3, CS; 126.4, C4"'; 126.6, C4"; 127.1, CS"; 129.0, C6"; Chapter 3 Experimental 159

129.1, C2"' and C6"'; 129.4, C3"' and CS"'; 130.5, C3"; 136.1, Cl"'; 136.7, Cl";

140.7, C2"; 175.8, C2. Mass spectrum: m/z 357 (M+ I, 2%), 356 (M+, 7), 214 (5), 200

(2), 169 (6), 155 (11), 143 (20), 129 (29), 123 (100), 121 (41), 115 (36), 91 (15), 77 (9),

65 (6), 45 (20).

Reactions of ester 137, lactone 139, acetal 149, sulfide 153 i. with gaseous HBr

(a) Ketone 138 (0.055 g, 0.166 mmol) was treated in dry

CH2Cl2 (1.0 mL) with gaseous HBr and the resulting solution kept in a sealed flask for 17 h at r.t. The solution was diluted with

CH2Cli (10 mL) and washed with water (2 x 10 mL), dried over 155 MgSO4 and evaporated to afford a brown oil (0.040 g) that was subjected to preparative t.l.c. using EtOAc-light petroleum (20:80) as developing sol vent to afford 4, 5 -trans-4-(2-methoxymethylphenyl)-5-methyl-2-propy/cyc/ohex-2- enone 155 as a pale yellow oil (0.017 g, 38%) (HRMS (ES): Found: mlz 295.1668.

C18H24O2Na requires: m/z 295.1668). IR Vmax (film): 2850, 2320, 1660, 1648, 1450,

1370, 1080, 750 cm·1• 1H n.m.r. 6 (CDC1 3, 300 MHz): 0.89, d, J 6.2 Hz, 5-Me; 0.90, t, J

7.7 Hz, H3'; 1.45, tq, J 7.7, 7.7 Hz, (H2')2; 2.20, m, 2H, CH2CH2CH3; 2.28, dm, J 7.5

Hz, H5; 2.26, m, H116; 2.59, d, J 12.0 Hz, Ha6; 3.38, s, OMe; 3.68, brdm, J 7.5 Hz, H4;

4.43, d, J 11.7 Hz, CHaHbOMe; 4.61, d, J 11.7 Hz, CHJ!bOMe; 6.49, dt, J 2.3, I.I Hz,

13 s:: H3; 7.14, dd, J7.5, 1.5 Hz, H3'; 7.21-7.36, m, 3xArH. C n.m.r. u (CDC13, 75.6 MHz):

13.9, C3'; 19.6, 5-Me; 21.7, C2'; 31.2, Cl'; 38.5, C5; 45.8, C4; 46.2, C6; 58.0, OMe;

73.0, ~H2OMe; 126.8, C4"; 128.0, C3"; 128.9, C5"; 129.9, C6"; 135.8, Cl"; 138.9, C2; Chapter 3 Experimental 160

142.1, C2"; 148.6, C3; 199.7 Cl. Mass spectrum: mlz 372 (M+, absent), 241 (M-31,

3%), 240 (18), 225 (5), 198 (33), 183 (18), 169 (47), 155 (46), 149 (100), 141 (62), 128

( 60), 115 (29), I 05 (8), 91 (8), 41 (20).

(b) A solution of ester 137 (0.059 g, 0.192 mmol) in dry 1 OMe

CH2Cl2 (1.0 mL) was saturated with gaseous HBr according to the procedure in (a). The usual workup gave a brown oil (0.057 1' 157 g) that was subjected to preparative t.l.c. by using CH2Cli-light petroleum (40:60) as developing solvent to give the major component methyl 3-[(l H)-isochromen-4- yl]butanoate 157 as a pale yellow gum (0.020 g, 46%) (HRMS (ES): Found: m/z

255.0989. C14H16O3Na+ requires: m/z 255.0991). IR Vmax (film): 2940-2830 (br), 1710,

1625, 1430, 1335, 1240, 1150, 1115, 1110, 1005, 940,850,750 cm-1• 1H n.m.r. o

(CDC13, 300 MHz): 1.23, d, J 5.3 Hz, (H4)J; 2.32, dd, J 15.1, 9.4 Hz, H8 2; 2.67, dd, J

15.1, 4.9 Hz, Hb2; 3.25-3.37, m, H3; 3.68, s, CO2Me; 4.90, d, J 12.4 Hz, Hal'; 4.96, d, J

12.5 Hz, Hbl '; 6.49, d, JI.I Hz, H3'; 7.03, dd, J 7.1 Hz, lxArH; 7.16-7.32, m, 3xArH.

13C n.m.r. o (CDC13, 75.6 MHz): 19.1, C4; 28.1, C3; 40.9, C2; 51.5, OMe; 68.1, Cl';

I 19.9, C3'; 120.0, C4"; 124.1, C5"; 126.6, C6"; 128.1, C3"; 129.0, Cl"; 130.6, C2";

173.0, Cl. Mass spectrum: m/z 233 (M+l, 5%), 232 (M+, 20), 217 (7), 188 (50), 173

(19), 159 (55), 149 (60), 145 (52), 131 (100), 119 (74), 115 (61), 105 (38), 91 (68), 77

(38), 69 (21), 51 (14), 41 (19). Chapter 3 Experimental 161

ii. with nitranium tetrajluarabarate (NO2BF,,J

Lactone 139 (0.049 g, 0.16 mmol) in dry CH2Cli (2.0 mL) was added dropwise cwfut to a well-stirred suspension ofNO2BF4 (0.024 g, 0.18 mmol) in dry CH2Cl2 (2.0 mL) in eu1 h.. ic;,-.!t~er Ar. The cooling bath was removed after 20 min, and stirring was continued at ambient temperature for another 1 h. The mixture was quenched with H2O (5.0 mL) and extracted with CH2Cli (3 x 10 mL), the extracts dried over MgSO4, and the solvent removed in vacua to afford a brown oil (0.038 g). 1H n.m.r. spectroscopic analysis revealed that the major components were unreacted lactone 139 and a trace of the desired aldehyde 159, but the aldehyde was unable to be isolated.

Acetal 149 (0.082 g, 0.269 mmol) in dry CH2Cl2 (1.0 mL) was added dropwise to a well-stirred suspension ofNO2BF4 (0.041 g, 0.30 mmol) in dry CH2Ch (1.0 mL) at

-23°C (CCliN2 bath) under Ar. The mixture was stirred for 45 min at -23°C and then placed in an ice-bath and stirred for another 1 h. The mixture was quenched with H2O

(5.0 mL) at -23°C, extracted with CH2Ch (3 x 10 mL), dried over MgSO4, and the

solvent removed in vacua to afford a brown oil (0.057 g). The oil was subjected to preparative t.l.c. on silica gel using EtOAc-CH2Cli (18:82) as solvent to afford in order

of increasing Rr value impure aldehyde 160 as a pale yellow oil (0.027 g, 35%) and

unreacted acetal 149 (0.029 g, 35%). Aldehyde 160 could not be further purified by

preparative t.l.c. on silica gel despite several attempts which led to gradual loss of

material. Chapter 3 Experimental 162

iii. with N-bromosuccinimide (NBS)

Freshly crystallized NBS (0.115 g, 0.644 mmol) was rapidly added to a solution of sulfide 153 (0.17 g, 0.478 mmol) in

CC14 (20 mL) containing benzoyl peroxide (0.023 g, 0.095 mmol).

The mixture was refluxed under white light for 3 h then cooled, 158 washed with water (20 mL), dried over MgSO4, and the solvent removed in vacuo to afford a brown oil (0.214 g). The oil was subjected to preparative t.l.c. on silica gel using EtOAc-light petroleum (45:55) as developing solvent to afford unreacted sulfide

153 (0.052 g, 31 % ) and ( 45)-{1-(2-formylphenyl)-2-phenylthioethyl]-(5S)-methyl-y butyro/actone 158 in 95% purity as a pale yellow gum (0.092 g, 57%) (HRMS (ES):

Found: 363.1029; C20H20O3SNa requires: 363.1025). IR Vmax (film): 3450, 2960, 2910,

1760, 1680, 1590, 1470, 1425, 1380, 1295, 1195, 1090, 935, 820, 735 cm·1• 1H n.m.r. o

(CDC13, 500 MHz, at 223K): 1.47, d, J 6.1 Hz, 5-CH3; 2.24, dd, J 18.1, 8.8 Hz, H113;

2.47, dd, J 18.1, 8.9 Hz, Ha3; 2.75, dddd,J 11.5, 8.9, 8.8, 8.6 Hz, H4; 3.20, dd, J 12.9,

8.9 Hz, CHaHbSPh; 3.34, dd, J 12.9, 4.8 Hz, CHaHbSPh; 4.67, dq, J 8.6, 6.1 Hz, H5;

4.54, ddd, J l 1.5, 9.0, 4.9 Hz, HI'; 7.17, dd, J 1.3, 7.1 Hz, H3"' and H5"'; 7.25, dd, J

7.3, 2.1 Hz, H2"' and H6"'; 7.40, dd, J 1.3, 7.1 Hz, H4"'; 7.51, m, H4" and H5"; 7.80, dd,J7.6, 1.0 Hz, H6"; 8.08, d,J7.l Hz, H3"; 9.95, s, CHO. 13C n.m.r. o(CDCl3, 125.7

MHz, at 223K): 22.0, CH3; 33.6, C3; 37.8, .CH2SPh; 39.0, Cl'; 46.1, C4; 80.6, CS;

126.6, C4"; 128.1, CS"; 128.4, C6"; 129.2, C2"' and C6"'; 129.3, C3"' and CS"'; 133.9,

Cl"; 134.4, C4"'; 135.0, C3"; 136.2, Cl"'; 142.0, C2"; 176.6, Cl; 194.5, CHO. Mass spectrum: mlz 341 (M+l, 15%), 340 (M+, 12), 279 (13), 231 (22), 223 (19), 205 (17), Chapter 3 Experimental 163

187 (23), 167 (49), 158 (74), 149 (100), 129 (85), 123 (95), 115 (86), 109 (46), 91 (43),

83 (28), 77 (41), 57 (48).

N.m.r. spectroscopic assignments were supported by HSQC, HMBC, DEPT 90 and 135, TOCSY, ROESY and NOESY experiments in CDC13 at 223K.

3.5. Ring Closure i. Reactions with lithium diisopropylamide (LDA)

General procedure i: n-BuLi was slowly added to a solution of diisopropylamine in dry THF at 0°C under Ar. The mixture was stirred for 10 min at this temperature and then cooled to -78°C for 15 min. A solution of the substrate, lactone 139 or acetal 149, respectively, in dry THF was added dropwise to the mixture and the mixture was stirred for another 15 min at -78°C whereupon its colour changed to orange colour. The reaction was allowed to continue for a measured amount of time, quenched and then worked up.

The general procedure i was adopted in the following cases:

(a) Lactone 139 (0.22 g, 0.72 mmol) in THF (4 mL) was added to n-BuLi (0.39 mL of 2.14 M, 0.83 mmol) and diisopropylamine (0.12 mL, 0.83 mmol) in dry THF (4.0 mL). The mixture was warmed to r.t., stirred for 1.75 hand quenched with water (8 mL) at r.t.. Extraction with EtOAc (3 x 20 mL) afforded a brown oil (0.13 g). The oil was subjected to preparative t.l.c. on silica gel using EtOAc-light petroleum (1:1) as developer. The major fraction comprised only unreacted lactone 139 (0.10 g, 44%). Chapter 3 Experimental 164

(b) Lactone 139 (0.082 g, 0.27 mmol) in THF (1.0 mL) was added to n-BuLi (0.15 mL of2.l M, 0.32 mmol) and diisopropylamine (0.046 mL, 0.32 mmol) in THF (2 mL).

The mixture was stirred at -78°C for 1 h, 3,4-dimethoxybenzaldehyde 163 (0.075, 0.45 mmol) in THF (1.0 mL) was added, stirring was continued for 1 h. The mixture was quenched with sat. NH4CI solution (5 mL), allowed to warm to r.t. then extracted with

EtOAc (3 x 20 mL) to afford a brown oil (0.14 g). None of the desired condensation product was detected by 1H n.m.r spectroscopic analysis. The oil was subjected to preparative t.1.c. on silica gel using EtOAc-light petroleum ( 1: 1) developer. The major fraction comprised only unreacted lactone 139(0.051 g, 62%).

(c) Lactone 139 (0.39 g, 1.28 mmol) in THF (2.0 mL) was added to n-BuLi (1.28 mL of 2.0 M, 2.56 mmol) and diisopropylamine (0.36 mL, 2.56 mmol) in THF (3 mL).

The mixture was warmed to r.t. then refluxed for 1.5 h and finally cooled to r.t., quenched with sat. NH4Cl (10 mL), extracted with Et2O (3 x 20 mL), and the extracts dried over MgSO4 and evaporated to dryness to afford a brown oil (0.22 g). The oil was subjected to preparative t.l.c. on silica gel using EtOAc-light petroleum ( 1: 1) as developing solvent. Two major fractions were collected but their 1H n.m.r. spectra both gave broad signals and no indication of the starting material 139 or desired product 161.

(d) Acetal 149 (0.056 g, 0.183 mmol) in THF (2.0 mL) was added to n-BuLi (0.13 mL of 1.62 M, 0.21 mmol) and diisopropylamine (0.03 mL, 0.21 mmol) in dry THF

(3.0 mL). The mixture was stirred at -78°C for I h, warmed to 0°C for I h, then cooled to -78°C again, and quenched with water (6.0 mL). The mixture was extracted with Chapter 3 Experimental 165

EtOAc (3 x 10 mL), the extracts dried over MgSO4 and evaporated to dryness to give a brown oil (0.068 g). None of the desired product 162 was detected by 1H n.m.r spectroscopic analysis. The oil was subjected to preparative t.1.c. on silica gel using

EtOAc-light petroleum (55:45) as developer to give as the major fraction, unreacted acetal 149 (0.045 g, 80%).

(e) Acetal 149 (0.054 g, 0.18 mmol) in THF (2.0 mL)) was added to n-BuLi (0.14 mL of 1.55 M, 0.21 mmol) and diisopropylamine (0.03 mL, 0.21 mmol) in dry THF

(3.0 mL). The mixture was stirred at -78°C for 2 h, warmed to r.t., stirred another 20 h, and then cooled to -5°C and quenched with water (5.0 mL). Extractive workup with

EtOAc gave an oil (0.025 g, 46%) that comprised acetal 156 but none of the desired cyclisation product 162 by 1H n.m.r. spectroscopic analysis.

(f) Acetal 149 (0.045 g, 0.15 mmol) in THF (1.5 mL) was added to n-BuLi (0.19 mL of 1.60 M, 0.30 mmol) and diisopropylamine (0.042 mL, 0.30 mmol) in dry THF

(1.5.0 mL). The mixture was stirred at-78°C for 1 h, warmed to r.t. for 1.75 h, recooled to -78°C, and quenched with water (5 mL). EtOAc extraction gave an oil (0.067 g, 76%) that contained acetal 149 as the major and only recognizable substance by 1H n.m.r. spectroscopic analysis.

ii. Reactions with KHMDS

General procedure ii: The substrates, ester 137, lactone 139 and acetal 149, respectively, in dry toluene was added to a solution of KHMDS in toluene at -90°C Chapter 3 Experimental 166

(N2/heptane) under Ar. The mixture was stirred for 1.5 hat -90°C, quenched with 3,4- dimethoxybenzaldehyde 163, or water, or TBDMSCl, and worked up.

General procedure ii was used in the following reactions:

(a) The reaction of ester 137 (0.10 g, 0.33 mmol) in THF (2.0 mL) and KHMDS

(0.98 mL of 0.5 M in toluene, 0.49 mmol) in THF (2.0 mL) was quenched with 3,4- dimethoxybenzaldehyde 163 (0.10 g, 0.60 mmol). The mixture was stirred for another 2 h, diluted with water and extracted with EtOAc (2 x 10 mL) to give a brown oil (0.184 g). None of the desired product 164 was detected by 1H n.m.r spectroscopic analysis.

The oil was subjected to preparative t.l.c. on silica gel using EtOAc-light petroleum

(40:60) as developing solvent. Two major fractions were collected as unreacted ester

137 (0.039 g, 39%) and aldehyde 163 (0.12 g).

(b) Reaction of ester 137 (0.10 g, 0.33 mmol) in toluene (1.5 mL) and KHMDS

(0.82 mL of 0.5 M in toluene, 0.41 mmol) in toluene (1.5 mL) was quenched with

TBDMSCl (0.10 g, 0.66 mmol). The mixure was stirred at -78°C for another 40 min and then at r.t. for 2 h. Evaporation of solvent under reduced pressure gave a brown oil (0.14 g) that by 1H n.m.r. spectroscopic analysis comprised unreacted ester 137 (0.05 g, 50%).

(c) The reaction of lactone 139 (0.200 g, 0.66 mmol) in dry toluene (4.0 mL) with KHMDS (1.5 mL of 0.5 M, 0.73 mmol) in toluene (4.0 mL) was stirred at -90°C for 15 min then quenched with a solution of 3,4-dimethoxybenzaldehyde OMe 163 (0.120 g, 0.73 mmol) in toluene (3.0 mL). Stirring was 166 Chapter 3 Experimental 167

continued for 1.5 h, the mixture quenched with H2O ( 10 mL) at -9CfC and the mixture allowed to warm to r.t. The mixture was extracted with EtOAc (3 x 20 mL), the aqueous layer was preserved, and the extracts were dried over Na2SO4. Solvent was removed in vacua from the organic extracts to afford a pale yellow oil (0.233 g). Preparative t.l.c. using EtOAc-light petroleum (70:30) as developing solvent gave recovered hydrazone

139 (0.072 g, 35%) and a 1: 1 mixture diastereomeric mixture of 3-[J-hydroxy-(3,4- dimethoxy)phenyl]methyl-(4R)-[J-(2-methoxymethyl)phenyl-2-(N,N dimethyl)hydrazono}ethyl-(5S)-methyl-y-butyrolactone 166 as a pale yellow, amorphous powder (0.081 g, 26%) m.p. 62-64°C (Found: C, 66.14; H, 7.23; N, 5.65. C26H34N2O6 requires: C, 66.34; H, 7 .28; N, 5.95%). IR Vmax (Nujol): 2930, 2870, 2560, 2270, 1760,

1590, 1500, 1450, 1430, 1370, 1250, 1180, 1075, 1025 cm·1• 1H n.m.r. o (CDCI3, 300

MHz): 1.21, 1.22, 2xd, J 6.7 Hz, 5-Me; 2.50-2.55, 2.60-2.65, 2xm, H4; 2.71, 2.72, 2xs,

NMe2; 2.83, dd, J 8.2, 3.6 Hz, H3; 3.34, s, CH2O0I3; 3.69, 3.72, s, 4"'-OMe; 3.75-3.83, m, HI"; 3.88, 3.89, s, 3"'-OMe; 4.17, dd, J 11.3, 2.0 Hz, CHaHbOMe; 4.38, dd, J 11.3,

5.1 Hz, CHJ!bOMe; 4.60-4.70, m, H5; 4.69, d, J 8.7 Hz, HI'; 4.85, brs, OH; 6.45, 6.46,

2xd, J7.7 Hz, H2"; 6.51, d, J 1.5 Hz, lxArH; 6.65, dd,J7.4, 7.7 Hz, lxArH; 6.70-6.81, m, 2xArH; 6.91-7.02, m, lxArH; 7.07-7.16, m, lxArH; 7.19-7.25, m, lxArH. 13C n.m.r. o (CDC1 3, 75.6 MHz): 21.1, 22.1, 5-Me; 42.9, 44.2, NMe2; 46.5, 47.3, Cl"; 52.1, 52.5,

C3; 55.6, 55.9, 3"'-OMe, 4"'-OMe; 58.1, CH2O,CH3; 72.6, 72.8, CH2OCH3; 72.9, 74.9,

Cl'; 80.6, 81.2, C5; 108.9, 109.3, C2"'; 110.4, 110.8, C4"'; 117.9, 119.4, C5"'; 126.5,

126.7, C4'"'; 128.1, 128.3, C5""; 128.3, 128.6, C6'"'; 129.9, 130.0, C3""; 132.7, 133.9,

Cl'"'; 135.6, 135.7, C2"''; 135.6, 135.7, C2"; 135.9, 138.5, Cl"'; 148.5, 149.0, C3"';

149.0, 149.2, C4"'; 178.7, 178.9, Cl. Mass spectrum: mlz 470 (M\ 3%), 452 (M-18, 3), Chapter 3 Experimental 168

304 (6), 273 (7), 228 (11 ), 205 (25), 191 ( 11 ), 173 (87), 167 ( 45), 139 (76), 130 (100),

115 (57), 95 (47), 77 (71), 59 (65), 43 (81).

The original aqueous solution was acidified with 30% H2SO4 (5.0 mL), and after

4h was extracted with EtOAc (3 x 20 mL ). Evaporation of solvent from the extracts gave aldehyde 148 as a pale yellow gum (0.067 g, 39%). The substance was identical in all aspects to that obtained from section 3.4.1, ii.

The yield of the alcohol hydrazone 166 was not changed when the reactants were added in the reverse order.

(d) The reaction was repeated under the conditions as described in ( c ), but the mixture was quenched with saturated aqueous NH4Cl solution rather than with aldehyde

163. Unfortunately, workup gave only unreacted lactone 139 (0.14g, 71%).

(e) Reaction of acetal 149 (0.050 g, 0.16 mmol) in dry toluene (1.0 mL) and

KHMDS (0.39 mL of0.5 M, 0.20mmol) in toluene (4.0 mL) was stirred at-90°C for 1 hand quenched with water (3 mL). EtOAc (3 x 10 mL) extraction and evaporation of solvent gave a brown oil (0.045 g, 89%) that by 1H n.m.r. spectroscopic analysis

comprised acetal 149.

(t) The reaction of acetal 149 (0.058 g, 0.19 mmol) in dry toluene (3.0 mL) with

KHMDS (0.42 mL of 0.5 M, 0.21 mmol) in toluene (4.0 mL), when quenched with

TBDMSCl (0.41 g, 2.7 mmol) in toluene (2.0 mL) and worked up by evaporation of Chapter 3 Experimental 169

solvent under reduced pressure gave a brown oil (0.068 g). 1H n.m.r. spectroscopic analysis showed the oil to comprise a complex mixture, but acetal 149 was not evident.

(g) The reaction of acetal 149 (0.045 g, 0.15 mmol) in dry toluene (3.0 mL) and

KHMDS (0.35 mL of 0.5 M, 0.21 mmol) in toluene (1.0 mL) was quenched with 3,4- dimethoxybenzaldehyde 163 (0.034 g, 0.21 mmol) in toluene (1.0 mL) and worked up by dilution in water and EtOAc extraction to give a brown oil (0.057 g). 1H n.m.r. spectroscopic analysis clearly showed unreacted acetal 149 (0.028 g, 62%), but none of the desired condensation product 167 was observed.

iii. Michael addition reaction of 117 with lactone 30 with trimethylsilyl chloride quench

N,N-Dimethylhydrazone 117 (0.67 g, 3.25 mmol) in dry THF (3.0 mL) was added dropwise to a stirred solution of n-BuLi (1.7 mL of 2.1 M, 3.57 mmol) in THF

(7.5 mL) at -78°C under Ar. The mixture was stirred for 15 min at -78°C whereupon its colour changed to deep red then Cul (0.12 g, 0.63 mmol) was added rapidly to the mixture and stirring was continued for another 10 min. Lactone 30 (0.32 g, 3.25 mmol) in THF (3.0 mL) was slowly added to the dark brown mixture at -78°C and stirring was continued for another 30 min. TMSCl (2.11 mL, 16.3 mmol) was added dropwise to the mixture, and the mixture was warmed to r.t. and evaporated to dryness in vacuo to afford a brown foam. The foam was dissolved in CH2Cli (20 mL), the solution filtered and the filtrate evaporated to give a brown oil (1.55 g).The brown oil was subjected to column chromatography on silica gel using EtOAc-light petroleum (1: 1) as developer to give a major fraction as a mixture of three diastereomers of (4S)-[1-(2- Chapter 3 Experimental 170

methoxymethyl}phenyl-2-(N,N-dimethylhydrazono)ethyl]-(5S)-methyl-y-butyrolactone

139 as a pale yellow oil (0.42 g, 42%). None of desired silyl enol ether product was observed.

iv. Reactions with Bu2BOTf/(iPr)iNEt

(a) n-Bu2BOTf (0.18 mL of 1.0 M, 0.180 mmol) was added to a solution of ester

137 (0.05 g, 0.16 mmol) in dry CH2Cl2 (2.0 mL) at 0°C under Ar. The mixture was stirred for 10 min and diisopropylethylamine (0.035 mL, 0.20 mmol) was added. The mixture was stirred at (fC for 30 min then cooled to -78°C and 3,4- dimethoxybenzaldehyde 163 (0.030 g, 0.18 mmol) in CH2Cli (1.5 mL) was added.

Stirring was continued at -78°C for 45 min, and then at r.t. for another 1.5 h. The mixture was diluted with CH2Cl2 (10 mL) and washed with H2O (3 x 10 mL), and the solvent evaporated to afford a brown oil (0.11 g) from which was isolated unreacted ester 137 as a pale oil (0.04 g, 80%). The identity of the substance was confirmed by 1H n.m.r. spectroscopic analysis.

(b) The reaction in (a) was repeated under the same conditions but without quenching with 3,4-dimethoxybenzaldehyde 163. It gave the unreacted starting material

137 (0.037 g, 75%).

(c) n-Bu2BOTf(0.160 mL of 1.0 Min CH 2Cl2, 0.160 mmol) was added to a solution of sulfide 153 (0.044 g, 0.123 mmol) in dry CH2Cli (2.5 mL) at -78°C under Ar. After the mixture was stirred for 10 min, diisopropylethylamine (0.03 mL, 0.185 mmol) was added to the mixture. The mixture was stirred at -78°C for 2 h then the contents warmed Chapter 3 Experimental 171

to 0°C for 30 min. BF3·OEt2 (0.023 mL, 0.185 mmol) was added to the mixture at -78°C.

Stirring was continued at this temperature for 1 h and then at 0°C for 1 h. The mixture was recooled to -78°C, quenched with H2O (3.0 mL), and allowed to warm to r.t .. The mixture was diluted with CH2Cl2 (10 mL) and the aqueous layer was extracted with

CH2Ch (3 x 10 mL). The combined extracts were dried over MgSO4 and evaporated to dryness to afford unreacted sulfide 153 as a pale yellow oil (0.034 g, 77%) by 1H n.m.r. spectroscopic analysis.

( d) The reaction in ( c) was repeated under the same conditions but without quenching with BF3·OEt2• It gave the unreacted starting material 153 (0.031 g, 70%).

(e) BF3·OEt2 (0.05 mL) was added to a solution of sulfide 153 (0.025 g) in dry

CH2Cl2 (1.5 mL) at r.t. The mixture was stirred for 3.5 days at r.t. and diluted with

CH2Ch and water (5 mL). The aq. solution was extracted with CH2Cl2 (2 x 10 mL) and the extracts evaporated to dryness to afford a semi-solid (0.019 g). 1 H n.m.r. spectroscopic analysis showed that there was neither significant starting material recovered nor desired product 169 detected.

v. Reactions ofaldehyde 158

(a) n-BuLi (0.076 mL of 1.04 M, 0.079 mmol) was added to a solution of diisopropylamine (0.012 mL, 0.082 mmol) in THF (5.0 mL) at 0°C under Ar. The mixture was stirred for 10 min at 0°C and then cooled to -78°C for 15 min. A solution of aldehyde 158 (0.023 g, 0.066 mmol) in dry THF (3.0 mL) was added dropwise to the mixture and it was stirred for 1.5 h at -78°C. The mixture was quenched with water (5 Chapter 3 Experimental 172

mL), extracted with EtOAc (3 x 15 mL), the extracts were dried over NaiSO4 and evaporated solvent in vacuo to afford a brown oil (0.046 g) that was not the starting material 158 by 1H n.m.r. spectroscopic analysis. The oil was subjected to preparative t.l.c. on silica gel using EtOAc-light petroleum (80:20) as developing solvent, but it was unsuccessful.

(b) Aldehyde 158 (0.070 g, 0.21 mmol) in toluene (2 mL) was added to a solution of K.HMDS (0.45 mL of 0.5 M, 0.23 mmol) in toluene (2 mL) at -90°C (N2/heptane) under Ar. The mixture was stirred at that temperature for 1.5 h and then quenched with water (4 mL). The mixture was warmed to r.t. and the phases separated and the aqueous layer extracted with EtOAc (3 x 15 mL). The combined extracts were dried over

MgSO4, then evaporated to dryness to afford a brown oil (0.06 g, 87%) which was unreacted aldehyde 158 by 1H n.m.r. spectroscopic analysis. 173

REFERENCES

1. H. Brodaty, R. Dresser, M. Eisner, T. Erkunjuntti, S. Gauthier, N. Graham, C.

Jonker, G. Sachs, and P. Whitehouse, Alzheimer Disease and Associated

Disorders, 1999, 13, 71.

2. J. Walton, and D. Bryson-Taylor, Chem. in Aust., 1995, 62, 37.

3. R. Tanzi, S. Gaston, A. Bush, D. Romano, W. Pettingell, J. Peppercorn, M.

Paradis, S. Gurubhagavatula, B. Jenkins, and W. Wasco, Genetica, 1993, 91,

225.

4. (a) D. C. Mash, D. D. Flynn, and L. T. Potter; Science, 1995, 228, 1115; (b) L.

L. Iversen; Trends Pharmacol. Sci., Suppl., 1986, 44; (c) M. McKinney, and T.

J. Coyle; Mayo Clinic Proceedings, 1991, 66, 1225.

5. J.C. Jaen, and R. E. Davis, Ann. Rev. Med. Chem., 1994, 29, 23.

6. T. Kubo, K. Fukuda, A. Mikami, A. Maeda, H. Takahashi, M. Mishian, T.

Haga, K. Haga, A. Ichiyama, K. Kangawa, M. Kojima, H. Matsuo, T. Hirose,

and S. Numa, Nature, 1986, 323, 411.

7. T. Kubo, K. Maeda, K. Sugimoto, I. Akiba, A. Mikami, H. Takahashi, T. Haga,

K. Haga, A. lchiyama, K. Kangawa, M. Kojima, H. Matsuo, T. Hirose, and S.

Numa, FEBS Lett., 1986, 209,367.

8. E. T. Peralta, A. Ashkenazi, J. W. Winslow, D. H. Smith, J. Ramachandran, and

D. J. Capon, EMBOJ., 1987, 6, 3923.

9. T. I. Bonner, N. J. Buckley, A. C. Young, and M. R. Brann, Science, 1987, 237,

527. References 174

10. T. I. Bonner, A. C. Young, M. R. Brann, and N. J. Buckley, Neuron, 1988, 1,

403.

11. (a) C. Melchiorre, Med. Res. Rev., 1990, 3, 327. (b) E. C. Hulme, N. J. M.

Birdsall, and N. J. Buckley, Ann. Rev. Pharmacol. Toxicol., 1990, 30, 633. (c)

M. P. Caulfield, Pharmacol. Ther., 1993, 58,319.

12. R. Hammer, C. P. Berrie, N. J.M. Birdsall, A. S. V. Burgen, and E. C. Hulme,

Nature, 1980, 283, 90.

13. C. Melchiorre, A. Cassinelli, and W. Quaglia, .I. Med. Chem., 1987, 30,201.

14. R. Hammer, E. Giraldo, G. B. Schiavi, E. Monferini, and H. Ladinsky, Life Sci.,

1986, 38, 1653.

15. S. A. H. Gilani, and L. B. L. Cobbin, Naunyn-Schmiedeb. Arch. Pharmacol.,

1986, 332, 16.

16. M. Li, R. P. Yasuda, S. J. Wall, A. Wellstein, and B. B. Wolfe, Mo/. Pharmacol.,

1991, 40, 28.

17. M. Waelbeoeck, M. Tastenoy, J. Camus, and J. Christopher, Mo/. Pharmacol.,

1990, 38,267.

18. W. Hoss, W. S. Messer, F. J. Monsma, M. D. Miller, B. R. Ellerbrock, T.

Scranton, S. Ghodsi-Hovsepian, M. A. Price, S. Balan, Z. Mazloum, and M.

Bohnett, Brain Res., 1990, 517, 195.

19. L. T. Potter, D. D. Flynn, H. E. Hanchett, D. L. Kalinoski, J. Lubemarood, and

D. C. Mash, Trends Pharmacol. Sci., 1984, 3 (Suppl.), 22.

20. R. Quirion, I. Aubert, P. A. Lapchak, R. P. Schaum, S. Teolis, S. Gauthier, and

D. M. Araujo, Trends Pharmacol. Sci., 1989, 10 (Suppl.), 80. References 175

21. M. H. Richards, Br. J. Pharmacol., 1990, 99, 753.

22. R. B. Barlow, K. J. Berry, P. A. M. Glenton, N. M. Nikolaou, and K. S. Sob, Br.

J. Pharmacol., 1976, 58,613.

23. G. Lambrecht, R. Feifel, B. Forth, C. Strohmann, R. Tacke, and E. Mutschler,

Eur. J. Pharmacol., 1988, 152, 193.

24. M. Waelbroeck, J. Camus, M. Tastenoy, E. Mutschler, C. Strohmann, R. Tacke,

G. Lambrecht, and J. Christophe, Eur. J. Pharmacol. Mo/. Pharmacol. Sect.,

1991, 206, 95.

25. A. Minarini, M. L. Bolognesi, R. Budriesi, M. Canossa, A. Chiarini, S.

Spampinato, and C. Melchiorre,J. Med Chem., 1994, 37, 3363.

26. P.A. Lapchak, D. M. Araujo, R, Quirion, and B. Collier, Brain Res., 1989, 496,

285.

27. K. Gulya, D. Budai, and P. Kasa, Neurochem. lnt., 1989, IS, 153.

28. J. W. Richard, A. Wilson, and R. Quirion, Soc. Neurosci. Abst., 1991, 17, 781.

29. H. N. Doods, R. Quirion, G. Mihm, W. Engel, K. Rudolf, M. Entzeroth, G. B.

Schiavi, H. Ladinsky, W. D. Bechtel, H. A. Ensinger, K. D. Mendia, and W.

Eberlein, Life Sci., 1993, 52,497.

30. F. J. Ehlert, W.R. Roeske, and H. I. Yamamura, Life Sci., 1994, SS, 2135.

31. C. J. Jaen and R. E. Recent, "Advances in the Design and Characterization of

Muscarinic Agonists and Antagonists", Chapter 3 in "Annual Reports in

Medicinal Chemistry", McCall Ed. (Academic Press: New York, 1994) pp. 23-

32. References 176

32. W. G. Eberlein, W. Engel, G. Mihm, K. Rudolf, B. Wetzel, M, Entrzeroth, N.

Mayer, and H. N. Doods, Trends Pharmacol. Sci., Suppl. IV, 1989, 50.

33. C. Melchiorre, A. Minarini, P. Angeli, D. Giardina, U. Goulini, and W. Quaglia,

Trends Pharmacol. Sci., Suppl. IV, 1989, 55.

34. R. M. Eglen, G. C. Harris, H. Cox, A. 0. Sullivan, E. Stefanich, and R. L.

Whiting, Naunyn-Schmiedeberg 's Arch. Pharmacol., 1992, 346, 144.

35. S. A. Darroch, W. C. Taylor, L. K. Choo and F. Mitchelson, Europ. J.

Pharmacol., 1990, 182, 131. .

36. X.-Q. Shi, MSc. Thesis, The University of New South Wales, 1996.

37. M. Kassiou, R. W. Read, and X.-Q. Shi, Bioorg. Med Chem. Lett., 1997, 7, 799.

38. R. F. C. Brown, R. Drummond, A. C. Fogerty, the late G. K. Hughes, J. T.

Pinhey, E. Ritchie, and W. C. Taylor, Aust. J. Chem. 1956, 9,283.

39. J. T. Pinhey, E. Ritchie, and W. C. Taylor, Aust. J. Chem., 1961, 14, 106. (c) E.

Ritchie and W. C. Taylor, "The Alkaloids IX", Ed. R. H. F. Manske, (Academic

Press: New York, 1967) pp. 529.

40. J. Fridrichsons and J.M. Mathieson, Acta Crystallogr., 1962, 15, 119.

41. M. Takadoi, T. Katoh, A. Ishiwata, and S. Terashima, Tetrahedron Lett., 1999,

40, 3399.

42. J. H. Miller, P. J. Aagaard, V. A. Gibson, and M. McKinney, J. Pharmacol. Exp.

Ther., 1992, 263,663.

43. A. P. Kozikowski, A. H. Fauq, J. H. Miller, and M. McKinney, Bioorg. Med

Chem. Lett., 1992, 2, 797. References 177

44. M. J. Malaska, A. H. Fauq, A. P. Kozikowski, P. J. Aagaard, and M.

McKinney, Bioorg. Med. Chem. Lett., 1993, 3, 1247.

45. M. J. Malaska, A. H. Fauq, A. P. Kozikowski, P. J. Aagaard, and M.

McKinney, Bioorg. Med. Chem. Lett., 1995, 5, 61.

46. J. Dunlap, and J. H. Brown, Mo/. Pharmaco/., 1983, 24, 15.

47. J. M. Stockton, N. J. M. Birdsall, A. S. V. Burgen, and E. C. Hulme, Mo/.

Pharmaco/., 1983, 23,551.

48. Ad. F. Roffel, C.R. S. Elzinga, H. Meurs, and J. Zaagsma, Europ. J. Pharmacol.

Mo/. Pharmaco/. Sect., 1989, 172, 61.

49. R. R. Levine and N. J. M. Birdsall, (Eds.) Trends Pharmaco/. Sci. 10 (Suppl.

Subtypes of Muscarinic Receptors IV}, VII, 1989.

50. A. L. Gardner, S. A. Darroch, L. K. Choo, and F. Mitchelson, Trends

Pharmacol. Sci., 9 (Suppl. Subtypes of Muscarinic Receptors III), 1988, 40.

51. K. Mori, Tetrahedron, 1975, 31,301 I.

52. M. Julia, and J.M. Paris, Tetrahedron Lett. 1973, 4833.

53. M. S. Gitler, R. C. Reba, V. I. Cohen, W. J. Rzeszotarski, and J. Baumgold,

Brain Res., 1992, 582,253.

54. D. J. Hart, W. ~L. Wu, and A. P. Kozikowski, J. Am. Chem. Soc., 1995, 117,

9369.

55. S. Chackalamannil, R. J. Davies, T. Asberom, D. Doller, and D. Leone; J. Am.

Chem. Soc., 1996, 118, 9812-9813.

56. D. J. Hart, J. Li, W. -L. Wu, and A. P. Kozikowski, J. Org. Chem., 1997, 62,

5023. References 178

57. J.E. Baldwin, R. Chesworth, J. S. Parker, and A. T. Russell, Tetrahedron Lett.,

1995, 36, 9551.

58. S. Hofman, G. De Baecke, B. Kenda, and P. J. DeClercq, Synthesis, 1998, 479.

59. D. Doller, S. Chackalamannil, M. Czarniecki, R. McQuade, and V. Ruperto,

Bioorg. Med. Chem. Lett., 1999, 9,901.

60. S. Chackalamannil, R. J. Davies, Y. Wang, T. Asberom, D. Doller, J. Wong, and

D. Leone, J. Org. Chem., 1999, 64, 1932.

61. G. De Baecke, and P. J. De Clercq, Tetrahedron Lett., 1995, 36, 7515.

62. (a) A. G. Fallis, Can. J. Chem., 1984, 62, 183. (b) G. Brieger, J. N. Bennett,

Chem. Rev., 1980, 80, 63. (c) E. Ciganek, Org. React., 1984, 32, I. (d) D. Craig,

Chem. Soc. Rev., 1987, 16, 187. (e) W. R. Roush, in "Comprehensive Organic

Synthesis", Eds. B. Trost, I. Fleming, (Pergamon Press: Oxford 1991) Chapt. 5,

pp. 513.

63. W. Eberiein, E. Engel, G. Trummlitz, G. Mihm, N. Mayer, and K. Hasselbach,

Trends Pharmaco/. Sci., 1988, Suppl. 92.

64. P. Perlmutter, Tetrahedron Org. Chem. Series, J. E. Baldwin, P. D. Magnus

(Eds.), "Conjugate Addition Reactions in Organic Synthesis", Vol. 9, pp. 283.

65. K. Iwai, M. Kawai, H. Kosugi, and H. UDA, Chem. Lett., 1974, 385.

66. R. M. Ortuno, D. Alonso, and J. Font, Tetrahedron Lett., 1986, 27, 1079.

67. S. -K. Kang, D. -H. Lee, and S. -G. Kim, Org. Prep. Proced. Int., 1990, 22,651.

68. T. Hiyama, K. Nishide and K. Kobayashi, Tetrahedron Lett., 1984, 2S, 569.

69. K. Omura, and D. Swem, Tetrahedron, 1978, 34, 1651.

70. R. D. Little, and G. W. Muller, J. Am. Chem. Soc., 1981, 103, 2744. References 179

71. T. Muraki and N. Mukaiyama, Chem. Lett., 1974, 1447.

72. T. Muraki and N. Mukaiyama, Chem. Lett., 1975, 215.

73. J. March, in "Advanced Organic Chemistry", 4th Ed., (Wiley-Interscience: New

York, 1992), pp. 795.

74. P. Perlmutter, Tetrahedron Org. Chem. Series, J. E. Baldwin, P. D. Magnus

(Eds.), "Conjugate Addition Reactions in Organic Synthesis", Vol. 9, pp. 45.

75. P. Perlmutter, Tetrahedron Org. Chem. Series, J. E. Baldwin, P. D. Magnus

(Eds.), "Conjugate Addition Reactions in Organic Synthesis", Vol. 9, pp. 33.

76. D. E. Bergeiter, and M. Momongan, in "Comprehensive Organic Synthesis", Vol.

2, C. H. Heathcock, Ed., (Pergamon Press: Oxford, 1991), pp. 503 and references

cited therein.

77. P. Perlmutter, Tetrahedron Org. Chem. Series, J. E. Baldwin, P. D. Magnus

(Eds.), "Conjugate Addition Reactions in Organic Synthesis", Vol. 9, pp. 46.

78. A. Sajtos, M. Wechsberg, E. Roithner, S. Pollhammur, and A. Mahringer,

European Patent 147593, 1985; Chem. Abstr., 1986, 104, 68447.

79. P. Garratt, and P. Vollhardt, Synthesis, 1971, 423.

80. E. Pokrorskaya, A. Menyailo, M. Pospelov, A. Ryshankova, A. Shil'nikova, G.

Dudnik, and L. Mishina, Neftekhimiya, 1970, 10, 554 (Chem. Abstr., 1970, 73,

120393).

81. S. Schaeren, and A. Furlenmeier, Fr. Patent 1545332, 1968 (Chem. Abstr., 1969,

71, 124817).

82. J. Gibietis, A. Zicmanis, I. Lipsbergs, and B. Spince, Latv. PSR Zinat. Akad.

Vestis, Kim. Ser. 1984, 233 (Chem. Abstr., 1984. 101, 30661). References 180

83. K. Potts, and R. Robinson, J. Chem. Soc., 1955, 2675.

84. P.H. Z. Hermans, Phys. Chem. Stoechiom. Verwandschaftsl., 1924, 113,365.

85. B. Wunsch, and M. Zott, Synthesis, 1992, 927.

85a U. Schmidt, A. Lieberknecht, H. Griesser, J. Talbiersky, J. Org. Chem., 1982, 47, 3261. 85b H. Itsuda, M. Kawamura, K. Kato, M. Sato, Jpn. Kokai Tokkyo Koho JP 62,155,236 (Cl. C07C47/565), 10 July 1987 (Chem. Abstr., 1987, 107, 217243u).

86a M. Rule, A.R. Martin, D.E. Seeger, E.F. Hilinski, D.A. Dougherty, J.A. Berson, Tetrahedron, 1982, 38, 787. 86b S. Murahashi, Scient. Papers inst. Phys. Chem. Res. (Tokyo), 1936, 30, 180.

89. I. Vogel, Practical Organic Chemistry, 3rd Ed., (Longmans: London, 1967), pp.

182.

90. A. Michael, J. Prakt. Chem., 1887, 3,349.

91. A. Michael, Am. Chem. J., 1887, 9, 112.

92. E. D. Bergmann, D. Ginsburg, and R. Pappo, Org. React., 1959, 10, 179.

93. F. A. Carey, and R. J. Sundberg, in "Advanced Organic Chemistry", 2nd Ed., Part

B, (Plenum Press: New York, 1983) pp. 31.

94. A. I. Meyers, A. Nabeya, H. W. Adickes, I. R. Politzer, G. R. Malone, A. C.

Kovelesky, R. L. Nolen, and R. C. Portnoy, J. Org. Chem., 1973, 38, 36.

95. E. D. Bergmann, D. Ginsberg, and R. Pappo, Org. React., 1959, 10, 179.

96. K. D. Croft, E. L. Ghisalberti, P. R. Jefferies, and A. D. Stuart, Aust. J. Chem.,

1979, 32, 2079.

97. K. Kobayashi, Y. Kajimura, K. Maed, T. Uneda, 0. Morikawa, and H. Konishi,

Heterocycles, 1997, 4S, 1593. References 181

98. D. Harrowven, and S. T. Dennison, Tetrahedron Lett., 1993, 34, 3323.

99. Y. Moritani, C. Fukushima, T. Ukita, T. Miyagishima, H. Ohmizu, and T.

Iwasaki, J. Org. Chem., 1996, 61, 6922.

100. M. Medarde, A. C. Ramos, E. Caballero, J. L. Lopez, R. P.-L. de Clairac, and A.

S. Feliciano, Tetrahedron Lett., 1996, 37, 2663.

101. K. Kobayashi, K. Maeda, T. Uneda, 0. Morikawa, and H. Konishi, J. Chem.

Soc., Perkin Trans. 1, 1997, 443.

102. H. Doucet, C. Bruneau, and P.H. Dixneuf, Synlett, 1997, 807.

103. S. Warren, in "Organic Synthesis: The Disconnection Approach", (J. Wiley &

Sons: Chichester, 1994), pp. 129.

104. G. B. Fisher, L. Lee, and F. W. Klettke, Synth. Commun., 1994, 24, 1541.

105. (a) G. Wilkinson, F. G. A. Stone, and E. W. Abel, in "Comprehensive

Organometallic Chemistry", Vol. 1, (Pergamon: Oxford, 1982), pp. 139. (b) J. F.

Garst, and R.H. Cox, J. Am. Chem. Soc., 1970, 92, 6389.

106. M. J. Gallagher, S. Harvey, C. L. Raston, and R. E. Sue, Chemical Commun.,

1988, 289.

107. J. March, in "Advanced Organic Chemistry", 4th Ed., (Wiley-Interscience: New

York, 1992), pp. 449.

108. M. F. Lappert, T. R. Martin, C. L. Raston, B. W. Skelton, and A. H. White, J.

Chem. Soc. Dalton Trans., 1982, 1959.

109. A. P. Kozikowski, Y.-P. Pang, F. Yamada, M. McKinney, and J. Miller, J. Med.

Chem., 1991, 34, 3399.

110. C. J. Li, Tetrahedron Lett., 1995, 36, 517. References 182

111. T. H. Chan and Y. C. Xin, Chem. Commun., 1996, 8,905.

112. Bruylants, and Feytmants-de Medicis, in Patai "The Chemistry of the Carbon

Nitrogen Double Bond", (Wiley: New York, 1970), pp. 465. J.A. 113. R. Jaques and .8". Leisten, J. Chem. Soc., 1964, 2683.

114. B. Bhatt, and L. Kulkarni, Synthesis, 1983, 249.

115. T.-L. Ho, and G. A. Olah, J. Org. Chem., 1977, 42, 3097.

116. (a) C. J. Easton, Chem. Rev., 1997, 97, 53. (b). R. Sustmann, and H. G. Korth,

Adv. Phys. Org. Chem., 1990, 26, 131.

117. P. J. Hore, in "Nuclear Magnetic Resonance", (Oxford Science Publications:

Oxford, 1995), pp. 45.

118. H. 0. House, W. V. Phillips, T. S. B. Sayer, and C.-C. Yau, J. Org. Chem., 1978,

43, 700.

119. (a). T. R. Williams, and L. M. Sirvio, J. Org. Chem., 1980, 45, 5082. (b). W. H.

Pirkle, and P. E. Adams, J. Org. Chem., 1990, 45, 4111.

120. D. A. Evans, J. Bartroli, and T. L. Shih,J. Am. Chem. Soc., 1981, 103, 2127. ,1so 121. L. Fieser, and M. Fieser, Reagents for Organic Synthesis,/jol. 8,427. tfJr 122. L. Fieser and M. Fieser, Reagents for Organic Synthesis, Vol. 6,498. /\

123. Claisen's alkali: A mixture of KOH (7.0 g), water (5 mL) and methanol (15 mL).

124. J. E. Cragg, R. B. Herbert, F. B. Jackson, C. J. Moody, and I. T. Nicolson, J.

Chem. Soc., Perkin Tans. I, 1982, 2477.

125. D. Murali, G. Rao and S. Krishna, Synthesis, 1987, 3,254.