Regiocontrol in the Heck-Reaction and Fast Fluorous Chemistry

Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy 244

______

KRISTOFER OLOFSSON

ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2001 Dissertation for the Degree of Doctor of Philosophy (Faculty of Pharmacy) in Organic Pharmaceutical Chemistry presented at Uppsala University in 2001.

Abstract

Olofsson, K. 2001. Regiocontrol in the Heck-Reaction and Fast Fluorous Chemistry. Acta Universitatis Upsaliensis. Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy 244. 74 pp. Uppsala. ISBN 91–554–4883–6

The palladium-catalysed Heck-reaction has been utilised in organic synthesis, where the introduction of aryl groups at the internal, β-carbon of different allylic substrates has been performed with high regioselectivity.

The β-stabilising effect of silicon enhances the regiocontrol in the internal arylation of allyltrimethylsilane, while a coordination between palladium and nitrogen induces very high regioselectivities in the arylation of N,N-dialkylallylamines and the Boc-protected allylamine, producing β-arylated arylethylamines, which are of interest for applications in medicinal chemistry. Phthalimido-protected allylamines are arylated with poor to moderate regioselectivity.

Single-mode microwave heating can reduce the reaction times of Heck-, Stille- and radical mediated reactions drastically from approximately 20 hours to a few minutes with, in the majority of cases, retained, high regioselectivity.

The use of heavily fluorinated tin reagents, which proved to be unreactive under thermal heating, is shown to be applicable with microwave-heating and the high fluorous content of the products is utilised with the aim of improving and simplifying the work- up procedure.

Kristofer Olofsson, Organic Pharmaceutical Chemistry, Department of Pharmaceutical Chemistry, Uppsala University, Box 574, SE-75123, Uppsala, Sweden

ISSN 0282–7484 ISBN 91–554–4883–6

Printed in Sweden by Lindbergs Grafiska HB, Uppsala 2000

2 List of Papers

The thesis is based on the following papers:

I. Kristofer Olofsson, Mats Larhed and Anders Hallberg Highly Regioselective Palladium-Catalyzed Internal Arylation of Allyltrimethylsilane with Aryl Triflates. J. Org. Chem. 1998, 63, 5076-5079.

II. Kristofer Olofsson, Sun-Young Kim, Mats Larhed, Dennis P. Curran and Anders Hallberg High-Speed, Highly Fluorous Organic Reactions. J. Org. Chem. 1999, 64, 4539-4541.

III. Kristofer Olofsson, Mats Larhed and Anders Hallberg Highly Regioselective Palladium-Catalyzed β-Arylation of N,N- Dialkylallylamines. J. Org. Chem. 2000, 65, 7235-7239.

IV. Kristofer Olofsson, Helena Sahlin, Mats Larhed and Anders Hallberg Regioselective Palladium-Catalysed Synthesis of β-Arylated Primary Allylamine Equivalents. J. Org. Chem. Accepted.

3 4 CONTENTS

LIST OF PAPERS...... 3

ABBREVIATIONS...... 6

1 INTRODUCTION...... 7

1.1 PALLADIUM(0) CATALYSED ORGANIC REACTIONS ...... 7 1.1.1 The Heck-Reaction ...... 8 1.1.2 The Stille-Reaction ...... 9 2 THE CATALYTIC CYCLE AND REACTANTS OF THE HECK-REACTION...... 10

2.1 THE CATALYTIC CYCLE...... 10 2.1.1 Catalyst Formation...... 10 2.1.2 Oxidative Addition...... 10 2.1.3 π-Complex Formation and σ-Complex Formation (Insertion) ...... 12 2.1.4 Elimination and Palladium(0) Regeneration ...... 17 2.2 REACTION CONDITIONS ...... 18 2.2.1 Palladium ...... 18 2.2.2 The Ligand (dppf) ...... 18 2.2.3 The Base ...... 21 2.2.4 The Solvent ...... 22 2.3 REGIOCONTROL IN THE HECK-REACTION ...... 23 2.3.1 The β-Effect of Allylsilanes...... 23 2.3.2 The Chelation of Palladium to Allylamines and Allylamides...... 26 3 RECENT DEVELOPMENTS...... 30

3.1 MICROWAVE CHEMISTRY ...... 30 3.1.1 The Microwave Oven...... 30 3.1.2 General Physics...... 31 3.1.3 Rate Enhancements with Microwave Chemistry...... 33 3.2 FLUOROUS CHEMISTRY...... 35 3.2.1 History ...... 35 3.2.2 Applications in Modern Chemistry ...... 35 4 AIMS OF THE PRESENT STUDY ...... 38

5 RESULTS AND DISCUSSION ...... 39

5.1 HIGHLY REGIOSELECTIVE INTERNAL ARYLATION OF ALLYLTRIMETHYLSILANE (PAPER I)...... 39 5.2 HIGHLY REGIOSELECTIVE INTERNAL ARYLATION OF N,N-DIALKYLALLYLAMINES AND PRIMARY ALLYLAMINE EQUIVALENTS (PAPER III AND IV)...... 45 5.3 INTERNAL ARYLATION OF FLUOROALLYLAMINES (APPENDIX) ...... 58 5.4 MICROWAVE HEATED FLUOROUS STILLE-COUPLINGS AND RADICAL REACTIONS (PAPER II)...... 60 6 CONCLUDING REMARKS ...... 64

7 APPENDIX...... 65

8 SUPPLEMENTARY MATERIAL...... 66

9 REFERENCES...... 70

10 ACKNOWLEDGEMENTS...... 74

5 Abbreviations

Ac acetyl AIBN azobisisobutyronitrile Ar aryl BINAP 2,2´-bis(diphenylphosphino)-1,1´-binaphthyl Boc tert-butoxycarbonyl BTF benzotrifluoride Bu butyl cbz carbobenzyloxy DABCO 1,4-diazabicyclo[2.2.2]octane disoppf (diisopropylphosphino)-ferrocene DMA dimethylacetamide DMF dimethylformamide dmpe bis(dimethyl-phosphino)ethane DMSO dimethylsulfoxide dppb 1,4-bis(diphenylphosphino)butane dppe 1,2-bis(diphenylphosphino)ethane dppf 1,1´-bis(diphenylphosphino)ferrocene dppm 1,1-bis(diphenylphosphino)methane dppp 1,3-bis(diphenylphosphino)propane EDG electron-donating group equiv equivalents Et ethyl EWG electron-withdrawing group FBS fluorous biphase system FC-84 a perfluorinated solvent FRP fluorous reverse phase GC gas chromatography L ligand MAO monoamine oxidase Me methyl MS mass spectroscopy NMR nuclear magnetic resonance NR no reaction Ph phenyl Pht (NPht) phthalimido PMP 1,2,2,6,6-pentamethylpiperidine ref reference SSAO semicarbazide-sensitive amine oxidase t tertiary TBAF tetrabutylammonium fluoride temp temperature Tf trifluoromethanesulfonyl TFA trifluoroacetic acid THF tetrahydrofuran TLC thin layer chromatography triflate trifluoromethanesulfonyl X halide or pseudohalide

6 1 Introduction

Palladium was discovered by William Hyde Wollaston in 1803 and was named after the asteroid Pallas, the second largest in the asteroid belt, that had been discovered the year before. The name Pallas itself can be traced to the Grecian goddess Pallas Athena. Palladium is nowadays produced mainly in Russia and South Africa and although this thesis will concentrate on the chemical applications of palladium is only 5% of the world’s palladium stock used in chemistry, while 50% is used in electric components and 10% for car-exhaust catalysts.

1.1 Palladium(0) Catalysed Organic Reactions

The ability to create new carbon-carbon bonds is of paramount importance in medicinal- and organic chemistry and there are today a number of methods available for that purpose. Palladium(0) catalysed organic reactions (e. g. Heck-,1-9 Stille-,10,11 and Suzuki-couplings12,13) belong to the most versatile and popular of these methods.

One explanation for the popularity of palladium chemistry is the wide array of functionalities that is compatible with the reaction conditions and the fact that the palladium metal, which is expensive, can be used in catalytic amounts. The Heck- coupling, which in most cases attaches a vinyl or aryl group to an olefin, is widely used today, particularly with electron-poor olefins, where the regiocontrol of the reaction generally is very good. However, some unsymmetrical olefins suffer from poor regioselectivity at the insertion step. The lack of regioselectivity can easily give a mixture of product regioisomers that can reduce yields and hamper work-up significantly (Scheme 1).

Y Y + X Pd + Y Y +

Scheme 1: Example of Low Regioselectivity in Heck-Reactions.

The development of methods where the regioselectivity of the Heck-coupling can be controlled should be of great importance both for the general understanding of the mechanism of the Heck-reaction as well as for broadening the applicability of the method.

7 1.1.1 The Heck-Reaction The palladium-catalysed coupling known as the Heck-reaction was first reported by Moritani,14 Fujiwara15 and Heck16,17 in the late 1960s. The reaction was stoichiometric at first but has been refined into a catalytic reaction. The standard model for the catalytic cycle, as described in Scheme 2, has recently been augmented with alternative mechanisms (e. g. Jutand’s new version of the catalytic cycle)6 and although these in many cases are valid only under special circumstances, there is today a growing awareness that the standard model for the Heck-reaction mechanism is not general. Pd(0)L4

Pd(0)L3

BaseHX RX Pd(0)L2 Pd(0) Regeneration Oxidative Addition Base

L H L R Pd(II) Pd(II) X L X L R L Y Y β-Elimination π-Complex Formation

L L L R Pd(II) R Pd(II) X X Y L Y Insertion

Scheme 2: The Classical Heck-Coupling Mechanism.

If the catalytic cycle is investigated in detail it is clear that one source of the confusion over the mechanism is the poor understanding of the individual steps. To gain knowledge thereof is a great challenge as there is an inherent difficulty in identifying the catalytically active complexes in a given reaction mechanism. The most stable structures, i. e. those that can be isolated and characterised, are in many cases not the conformations responsible for the success of the reaction.6 The difficulty of coming to terms with the details of the catalytic cycle is highlighted by the comparatively large

8 number of papers on the Heck-reaction mechanism that still, approximately thirty years after the first report of the coupling, are published in respected journals.

1.1.2 The Stille-Reaction The Stille-reaction is a cross-coupling reaction where an alkynyl, vinyl, aryl or alkyl group is transferred from a tin reagent to an organohalide (Scheme 3).

Pd(II)

Reduction

3 Pd(0)Ln R -X R1-R3

Reductive Oxidative Addition Elimination

R3 R3 LnPd(II) Ln Pd(II) R1 X

Transmetallation

2 1 2 X-Sn(R )3 R -Sn(R )3

Scheme 3: The Mechanism of the Stille-Reaction.

The stability of the organotin starting materials makes the Stille-reaction a popular and generally very reliable method for carbon-carbon bond formation. The toxicity of the tin compounds and the relative sluggishness of the method, where reaction temperatures of up to 100 °C are not uncommon, reduce the synthetic usefulness of the coupling.18 LiCl is frequently added to stabilise the palladium catalyst10 and Cu(I) salts have also been utilised to improve the selectivity at the rate-determining transmetallation step. Copper salts may function both as phosphine scavengers, when PPh3 is used as ligand, and as a means of converting the stannane to an organocopper species.18

9 2 The Catalytic Cycle and Reactants of the Heck-Reaction

2.1 The Catalytic Cycle

2.1.1 Catalyst Formation Palladium is usually administered in its Pd(II) oxidation state and is reduced in the reaction medium to the catalytically active Pd(0) form (Scheme 2). There are at least four possible reducing agents in the typical Heck-coupling: the phosphine ligand,19-21 the base,22 the olefin23,24 or, in some cases, the solvent.25 There has been some uncertainty over which reducing agent predominates over the other, but convincing evidence has emerged that phosphine ligands, when they are present, are the most active of the four and the agent most likely to reduce Pd(II).26 The mechanism depicted in Scheme 4 has been proposed for the reduction of Pd(II) species.20

AcO OAc AcO Pd(II)(OAc)2 + 2 PPh3 Pd(II) Pd(0) + AcO-PPh3 Ph3PPPh3 Ph3P 1

AcO PPh3 +AcO O Ac2O + O=PPh3

O AcO-PPh3 Me

+H O 2 AcO O AcOH + O=PPh P H 3 -H Ph3

Scheme 4: Reduction of Pd(II) to Pd(0) by Phosphine Ligands.

2.1.2 Oxidative Addition In the oxidative addition step is Pd(0) oxidised to Pd(II) and an aryl- or vinyl group is added with its X–-counterpart, which usually is a halide or a pseudohalide (Scheme 5). Examples of pseudohalides that have been used in Heck-reactions include triflates,27 nonaflates,28 aryldiazonium salts7 and iodonium salts.7 L X Pd(0)L2 + ArX Pd(II) 2 Ar L

Scheme 5: Oxidative Addition of Palladium(0) to ArX.

The rate of the oxidative addition29 is quite sensitive to the choice of halide or – – – – pseudohalide with the reactivity falling in the series ArN2X > I >> OTf > Br >> Cl . Reactions with aryldiazonium salts are not very common but aryl- and vinyl iodides

10 have been studied extensively. Iodides are usually reactive at room temperature even without the addition of phosphine ligands. Aryl- or vinyl triflates, nonaflates and bromides typically need phosphine ligands, if the reactions are not run under phase transfer conditions,5 but are usually quite reactive. It is interesting to note that selectivity between aryl triflates and -bromides has been attained under certain reaction conditions.30 Aryl chlorides react only with difficulty and at elevated temperatures in standard Heck-couplings and considerable effort is at present made to explore this reaction.31-36 One incentive for this exertion is the relative cheapness and availability of aryl chlorides, when compared to aryl triflates and the other aryl halogens, which make the chloro-compounds attractive for industrial-scale synthesis. The development of water-soluble catalysts7 is another research area that is highly interesting for industrial applications.

The choice of organic moiety that is to be coupled (–R, Scheme 2) is limited as it should not have a cis-hydrogen on an sp3-hybridised carbon in the position β to the palladium. The desired reaction may be ineffective if this prerequisite is not fulfilled since Pd–H elimination of the cis-hydrogen occurs more readily than addition to olefins.1 Suitable – R groups can be e. g. aryl,1 alkenyl,1 benzyl,1 allyl,37 alkynyl37 or alkoxycarbonyl- methyl37 groups.

Although the 14 e--complex 2 (Scheme 5) is commonly referred to as the reactive species in the oxidative addition of monodentate phosphine-ligated palladium - - complexes has the 16 e -complex Pd(0)(PPh3)2(OAc) 3 (Scheme 6) recently been presented as a contender.6 Complex 3 is formed from complex 1 (Scheme 4) by equilibrium with one equivalent of phosphine ligand. The oxidative addition may then proceed through a short-lived pentacoordinated palladium species 4, yielding 5, the product of the oxidative addition (Scheme 6).

PPh3 PhI I -I Pd(PPh ) (OAc) Ph Pd PhPd(PPh ) (OAc) 3 2 OAc 3 2 3 PPh3 4 5

Scheme 6: Oxidative Addition via a Pentacoordinated Palladium Species.

The oxidative addition of bidentate-ligated palladium complexes has recently been reported to go through the full dissociation of one of the chelated bidentate ligands (Path A, Scheme 7)38 or alternatively through the reversible dissociation of one of the arms of one bidentate ligand in consequence of the association of ArX through a combination of associative and dissociative steps (Path B, Scheme 7), finally generating the same Pd(0) complex 6 as Path A. These mechanisms were found to be applicable to BINAP as well

11 as dppf.38 Notable is that one equivalent of bidentate ligand is in both mechanisms ligated to the palladium through all intermediate steps of the oxidative addition.

Pd(0) Pd(0) Pd(II)

L L L Ar L ArX A: Pd Pd + LL Pd L L L L X 6 L L L ArX L Ar B: Pd L Pd Pd Pd L L L LL L Ar-X L X 6

Scheme 7: Mechanism for Oxidative Addition with Bidentate Ligands.

2.1.3 π-Complex Formation and σ-Complex Formation (Insertion) It is vital to control the π-complex formation and insertion39,40 steps in order to control the regioselectivity of the Heck-reaction, that is if the –R group will be added to the internal carbon of the olefin, yielding 7, or the terminal, yielding 8 (Scheme 8).

π-Complex Formation Insertion L R L L Pd(II) +L Pd(II) R L R X X Y Y RY Pd(II) Y 7 X L -L L R L L R Pd(II) +L Pd(II) Y X X R Y 8 Y Scheme 8: Internal and Terminal Insertion in the Heck-Reaction.

The literature has described two slightly different mechanisms of the Heck-reaction, depending on if the leaving group is a tightly coordinating halide, typically bromides (the neutral pathway), or a weakly coordinating pseudohalide, typically triflates (the cationic pathway). There are a number of reaction parameters that can be chosen to ensure that only one, or mostly one, of the pathways is active in a certain reaction. If the reaction parameters are set ambiguously can both the neutral and cationic pathway be active simultaneously and mixtures of regioisomers may be produced if unsymmetrically substituted olefins are used. Some previously reported regioselectivities for the arylation of vinyl- and allyl substrates are summarised in Scheme 9.3

12 Ratio Internal/Terminal Ratio Internal/Terminal Insertion Insertion

Pathway Pathway Entry Olefin Neutral Cationic Entry Olefin Neutral Cationic

OH 1. COOMe 0/100 0/100 5. 0/100 100/0 O OH 2. N 40/60 100/0 6. 10/90 95/5

3. OBu Mixture of 100/0 7. OH 20-25/ 90/10 Isomers 80-85

4. OAc Mixture of 95/5 8. 20/80 80-85/ Isomers 20-25

Scheme 9: Regioselectivity in the Neutral and Cationic Heck-Reaction.

The neutral pathway: The choice of an aryl- or vinyl halide as substrate results in a neutral oxidative addition complex, e. g. 9 (Scheme 10), where the anion (–X) is tightly coordinated to the palladium. When the olefin coordinates to the complex is one of the phosphine ligands lost rather than the halide –X, which is why monodentate ligands are used in neutral-pathway Heck-reactions; they coordinate more weakly to palladium than bidentate ligands.

LX LL Pd(II) Pd(II) Ar Ar 9 10 R R Scheme 10: Neutral and Cationic π-Complexes.

Factors governing the regiochemical outcome of Heck-reactions going through the neutral pathway are largely sterical. The neutral complex 9 will preferably give the linear, terminally arylated product 8 (Scheme 8) as the organic moiety (the –R group) usually is larger than the palladium moiety and will be inserted on the less substituted carbon.1 When electron-withdrawing substituents are present on the olefin is the electronic effect inducing the insertion to the same position as the steric effect, that is stimulating terminal substitution (e. g. entry 1, Scheme 9). With electron-rich groups present on the olefin, may the electronic factors favour insertion at the carbon with the highest electron density (the internal, secondary carbon in the case of electron-rich, terminal alkenes) instead of the carbon with the lowest degree of substitution (the terminal carbon of the olefin), which may give a mixture of products (entries 3–4,

13 Scheme 9). Olefins with electron-withdrawing groups are thus favoured in reactions going through the neutral mechanism.41

Even though the electronic influences of the reactants generally are lower in the neutral pathway than in the cationic has an unmistakable change in regioselectivity between electron-poor and electron-rich or -neutral aryl halides been reported in the arylation of enol ethers (Scheme 11).42 Butyl vinyl ether that easily gives a mixture of regioisomers could with a relatively high regioselectivity be arylated at the terminal carbon with p- nitrophenyl halides.42

Br Br

O2N

23 62

77 38 OBu OBu

Scheme 11: Ratio of Arylation of Butyl Vinyl Ether with Electron-Poor and -Neutral Aryl Bromides.

All in all, the electronic properties of the olefin are of less importance in the neutral pathway than in the cationic, as the neutral complex is associated with a reduced polarisation of the double bond, but can in some instances influence the regiochemistry of the insertion.3,43

The cationic pathway: The most popular pseudohalide in Heck-reactions going through the cationic pathway is the triflate. A high selectivity for internal arylation of butyl vinyl ether in coordinating solvents was reported by our group in the late eighties.44 Cabri has since then published several seminal papers on the use of aryl triflates in combination with palladium catalysts ligated to bidentate ligands in reactions with electron-rich olefins.3 Triflate groups will, especially in polar solvents such as DMF,45 easily dissociate from the oxidative addition complex and produce a cationic complex 10 (Scheme 10). As the triflate anion is to be exchanged upon coordination to the olefin are bidentate ligands such as dppf or dppp used, which are not as easily exchanged as monodentate ligands and thus indirectly favour the dissociation of the triflate group.

The electronic properties of the olefin are in the cationic pathway important not only for the reactivity but also, to a higher degree than the neutral pathway, for the regiochemical outcome of the insertion. The coordination of the olefin in a cationic complex increases the polarisation of the π-system of the olefin, easing the migration of the aryl moiety onto the internal carbon because of the lower charge density with respect to the terminal carbon.43 This is not to say that steric factors are to be ignored in

14 the cationic mechanism. In fact, several studies indicate that there is, in many cases, a balance between electronic and steric factors that influences the regioselectivity in reactions going through the cationic pathway;7,41 a manifestation that adds to the difficulty of finding, or expressing, a simple model for the mechanism of the Heck- reaction.

The mechanism of the cationic Heck-reaction is described in Scheme 12. Electron-rich olefins generally function better than electron-poor olefins.3 One factor behind this phenomenon may be the disfavoured interaction between an electron-poor olefin and the cationic palladium complex,41 as depicted in complex 10, Scheme 12, leading to a slow insertion. The insertion step has been suggested to be the rate-determining step in the cationic Heck-mechanism.6,43

BaseHOTf L L ArOTf Pd(0) Regeneration Pd(0) Oxidative Addition Base

L L L L Pd(II) Pd(II) H OTf Ar OTf

Ar R OTf OTf R β-Elimination π-Complex Formation

L L L L Pd(II) Pd(II) Insertion Ar Ar R 10 R

Scheme 12: The Cationic Heck-Reaction Mechanism.

The reaction (and the regioselectivity) can in many cases, when vinyl- or aryl halides are used, be changed from the neutral- to the cationic pathway by adding halide “scavengers”, such as silver46-48 and thallium43,49 salts (Scheme 13).

L X Ag- or Tl-salts L Pd(II) Pd(II) + AgX or TlX Ar L Ar L

Scheme 13: Scavenging of Halide Anions by Silver- or Thallium-Salts.

15 There may also be other connections between the neutral and cationic pathways as Jutand has reported the dissociation of the halide in neutral palladium complexes 11 in coordinating solvents such as DMF, yielding cationic complexes 12 (Scheme 14).50

ArPdL2X ArPdL2(DMF) 11 12

Scheme 14: Equilibrium Between Neutral and Cationic Arylpalladium Complexes in DMF.

Overman has studied the insertion of bidentate-ligated neutral- as well as cationic palladium species and found proof that a temporary dissociation of one of the arms of BINAP is unlikely, even in the neutral pathway, as this should have a negative effect on the stereoinduction; a result that was not seen. A mechanism involving a transient neutral, pentavalent palladium complex 13 was proposed with aryl halides in the neutral mechanism, while a cationic complex 14 could be formed directly from aryl triflates (Scheme 15).51

Neutral LL LL L Ar ArX Pd Pd Pd -X XAr L X 13

LL LL Pd X Pd Ar X Ar

Cationic LL LL LL ArOTf Pd Pd Pd Ar solv OTf 14 OTf Ar

Scheme 15: Neutral and Cationic Pathways with Palladium-BINAP Complexes.

16 2.1.4 β-Elimination and Palladium(0) Regeneration The β-hydride elimination is the step yielding the final product of the Heck-reaction. For this process to occur must the insertion complex be able to rotate to a position where a β-hydrogen is placed cis to the palladium. The elimination will then result in a reconstituted alkene and a palladium hydride species. The β-elimination is reversible and the preferred formation of the thermodynamically more stable trans products is thus explained (Scheme 16).37

PdH Pd H -PdH H H Y R YR YR Pd R H H Y H H Pd H Pd R -PdH R H R Y H Y Y Scheme 16: β-Elimination in the Heck-Reaction.

There is today no precise knowledge of how Pd(II) is reduced to Pd(0). A recently published mechanism favouring a base-promoted elimination of Pd(II) to Pd(0) (Scheme 17) is suggested to be more energetically favoured than a mechanism where a 52 β-elimination yields a free HPdL2 species.

R´3N R´3NHX PdL HPdL2 X 2 PdL2 + R R R

Scheme 17: Base-Promoted Elimination.

17 2.2 Reaction Conditions

Evidently, it is difficult to analyse the precise importance of every ingredient of a reaction and the Heck-reaction has proven to be a tough nut to crack, as has been noted not without humour in a recent review by Beletskaya,7 where it is amply demonstrated that even small variations in the reaction conditions can have considerable effects on the overall performance of the system. Some of the components37 of the Heck-reaction are reviewed below.

2.2.1 Palladium There is at present an abundance of different palladium complexes available.37 It is difficult to predict how a certain catalytic system will work under given reaction conditions and it is not uncommon to test several different catalyst/ligand pairs when a new reaction is to be optimised. However, mechanistic investigations have suggested a role for the –OAc anion in the catalytic cycle6,19 and the relatively moisture insensitive

Pd(OAc)2 catalyst is perhaps the first choice today.

2.2.2 The Ligand (dppf) The choice of ligand can have a considerable effect on the outcome and selectivity of the Heck-coupling. Phosphine ligands are regularly used as they stabilise reactive intermediates and influence the reactivity and selectivity through electronic and steric effects.7,53,54 The presence of phosphine ligands can reduce the rate of readdition, and increase the rate of dissociation, of the hydridopalladium to the double bond in a Pd– olefin π-complex intermediate.1 This would tend to decrease double-bond migration and the formation of unwanted β-elimination products.

One of the most commonly used bidentate ligands is 1,1´-bis(diphenylphosphino)- ferrocene (dppf).55,56 Dppf is a sandwich-complex and is known to be flexible in three- dimensional space,55,57 probably more so than the other most ordinarily used bidentate ligands, such as dppp and BINAP. Several studies support the idea that ligands with a large bite angle54,58 (Scheme 18) can increase the rate of palladium elimination, as the large bite angle induces a stress in the structure of the intermediate complex that will be released by the elimination.26,54,55,59,60

Bridge θ LL Pd

Scheme 18: Representation of the Bite Angle (θ) of a Bidentate Ligand.

18 Analogues of dppf that have an even larger bite angle, e. g. (diisopropylphosphino)- ferrocenes (disoppf), are known to give very high regioselectivities. Too large bite angles, on the other hand, are known to be deleterious for the Heck-reaction.54

LIGAND BITE ANGLE dppp54 91° BINAP61 92.7° dppf56 99.1°

disoppf60 103.4°

If dppf is added in more than two equivalents to the amount of palladium can the complex 15 be formed (Scheme 19). It was previously assumed that a dissociation of one of the pairs of the bidentate ligands was necessary in order for a Heck-reaction to take place62 but recent results, where the complex 15 was found to have catalytic activity with aryl halides,38 put doubt on this suggestion. Instead, it has been suggested that 15 might be the dominating resting state of the active catalyst in the reaction medium.38,63

LL LL Pd(0) Pd(0) + LL LL 15 Scheme 19: The Equilibrium of PdL4 and PdL2 with Bidentate Ligands.

Another explanation for the success of bidentate ligands in catalysis might be formation of the palladacyclic intermediate 16, which was identified after 6 h at 100 °C in DMA (Scheme 20).38

Br PPh2 Br PPh2 Pd(dppf)2 + Fe Pd Fe Br P Pd Ph2 PPh 16 3

Scheme 20: Formation of Palladacycle Intermediates with Pd-dppf Complexes.

The importance of palladacyclic complexes in palladium-catalysed reactions with bidentate ligands is at present poorly understood but palladacycles are known for their stability64 and the formation of 16 could protect the catalytic system from decomposition at high reaction temperatures and during prolonged reactions.

19 The electronic influences of dppf on palladium are somewhat unclear. Electrochemical studies of iridium complexes indicate that dppf is at least a stronger electron donor than dppm or dppe.65 The possibility of dative bonding between the iron centre of dppf and palladium has been implied as well.55,66

A side reaction that can be seen in some Heck-couplings is the generation of olefins arylated with an aromatic ring from the ligand system. This process is also called aryl migration or aryl scrambling42,64,67 and stems from the proximity of the palladium- bound aryl group and the aryl groups of the ligand in complexes such as 17 (Scheme 21). This proximity might result in the insertion of one of the aryl groups from the ligand instead of the desired aryl group. In Scheme 21 have the aryl groups in the dppf- complex been pictured as lying flat in the dimension of the paper for the sake of an easier overview. MeO Y

OTf Y Pd Pd + P P P P MeO -OTf Fe Fe

MeO

Pd Y P P +

Scheme 21: Aryl Migration in the Pd/dppf Complex.

Palladium complexes with electron-donating ligands seem more prone to aryl exchange than those with electron-withdrawing ligands.67-69 The migration of alkyl groups69 has also been noted, although the mechanism may be different than that for aryl migration, which has been elucidated by Novak.67 The mechanism of the aryl migration may involve the dissociation of a phosphine from the palladium centre as the addition of 1 67 equivalent of PPh3 in mechanistic studies suppresses the scrambling almost totally, 70 while 0.1 equivalent of PPh3 did not influence the reaction at all. This could be explained by a mechanism encompassing reductive elimination of the Pd(II)L2Ar´X 67,71,72 species to Pd(0) (Scheme 22). The addition of excess PPh3 could lead to the formation of Pd(PPh3)n from Pd(0), preventing subsequent oxidative addition.

20 Ar' Ar3P X Ar2P X Pd(II) Ar'PAr3 + X + Pd(0) Pd(II) Ar' PAr3 Ar PAr3 L

Pd(0)Ln

Scheme 22: Aryl Migration by Reductive Elimination.

Recent studies indicate that aryl migration is dependent both on the nature of X in ArX,73 where the ease of scrambling is I- > Br- > Cl-, and the nature of the solvent, with highly polar solvents giving more scrambling than less polar. These results would support the theory that Pd–X ionisation is a requirement for the exchange to occur.67,73 As many studies in which aryl migration is investigated are conducted with GC–MS it is worth noting that no aryl exchange seem to occur during the process of mass measurement.70

2.2.3 The Base The role of the base in the Heck-coupling has been debated and although some clarity has been added, it is nonetheless safe to say that our knowledge of the details of the functions of bases in the reaction medium is unsatisfactory. It was suggested from an early point that bases were needed to deprotonate the hydridopalladium complex formed at the β-elimination step in order to regenerate Pd(0) (Scheme 2). This mechanism has been corroborated recently.74 Jutand has suggested that nitrogen bases may be involved in the abstraction of protons that otherwise would interact with the –OAc anion in complex 3 (Scheme 6), and produce the proton-coordinated complex 18 (Scheme 23). Complex 18, where the acetate is more weakly coordinated to the palladium than in 3 would result in a more naked and reactive palladium species.6 The proton abstraction would occur less readily when bases are added and bases may thus slow the oxidative addition.

Ph3P Pd(0) OAc H Ph P 3 18 Scheme 23: Abstraction of Palladium-Ligated Acetate by Protons in the Reaction Medium.

Inorganic bases have in several papers been used successfully in couplings where a lower degree of reduction of the aryl precursors was needed.75 This could be explained, when aryl bromides are used, by the neutralisation of the HBr formed during the reaction.76 Nitrogen bases have also been suggested to act as a source of hydrides in the deoxygenation of triflates, which would explain the low degree of deoxygenation when inorganic bases are used in favour of organic (Scheme 24).75,77

21 TfO PdL2Ar H H PdL2Ar Et2N H NEt2 H PdL2HAr + NEt H3C H 2 TfOH Scheme 24: Organic Bases as a Source of Hydrides.

The base can also affect an equilibrium that influences the concentration of the key reactive species of the insertion, ArPd(II)(OAc)(PPh3)2 5, as illustrated in Scheme 25. The abstraction of hydrogens is thought to shift the equilibrium to favour the reactive 5 + 6 over the less reactive cationic complex ArPd(II)(PPh3)2 19.

ArPd(OAc)(PPh3)2 ArPd(PPh3)2 + OAc 519 HNEt3

HOAc Scheme 25: Equilibrium Shift due to Abstraction of Hydrides by a Base.

The base has thus at least two roles; one that impedes the fast oxidative addition (Scheme 23) and one that speeds the slow π-coordination and insertion (Scheme 25). This is suggested to be beneficial for the catalytic cycle as the oxidative addition and insertion steps then can take place with comparable rates.6

2.2.4 The Solvent Polar, aprotic solvents such as acetonitrile and DMF are often used in Heck reactions.37 The possible effects of trace amounts of water in the reaction medium have been discussed, with varying conclusions. Brown indicated that just traces of water could alter the composition of the palladium complexes in reactions with bis(phosphane)- palladium complexes,74 while Overman found no noticeable effect of up to 5% (v/v) water in couplings with Pd/BINAP systems.51 Poorly degassed reaction solutions were on the other hand clearly shown to have a negative effect on the stereoselectivity due to oxidation of phosphine ligands to phosphine oxides by oxygen.51 Jutand has noted that the presence of water does not affect the rate of reduction of Pd(II) to Pd(0).20 Crisp suggested that water might be a source for hydroxide ligands that in reactions with aryl chlorides temporarily can replace the halogen and produce a more reactive species.78 Traces of water may also increase the degree of aryl scrambling (especially when reactions are run with relatively unpolar solvents) by making the reaction medium more polar.67

22 2.3 Regiocontrol in the Heck-Reaction

2.3.1 The β-Effect of Allylsilanes The carbon–silicon bond is in a number of ways different from the carbon–carbon bond and these differences are reflected in the unique chemistry of organosilicon compounds.79-81 Silicon is one of the few elements routinely used in synthetic organic chemistry that are more electropositive than carbon (Pauling electronegativity for Si: 1.8 and C: 2.5)82 and the C–Si bond is also longer (typically 25%)82 than the C–C counterpart and relatively easy to polarise.83 Silicon also forms stronger bonds with electronegative elements than with carbon and the addition of fluoride salts such as TBAF can often be used for the removal of silicon groups. Electrophilic substitution of the silyl functionality has also frequently been taken advantage of in preparative chemistry.79,84,85 Another peculiarity of silicon that can be put to synthetic use is the ability of organosilicon compounds to stabilise a cation in a β-position, the so-called β- effect of silicon (Scheme 26).

SiMe β 3 γ α

Scheme 26: Denomination of the α, β and γ -Positions.

The basis of the β-effect is thought to be the σ–π hyperconjugation (Scheme 27).86-88 Another form, the p–d homoconjugation87 (Scheme 27) is also discussed in the literature and uncertainty in the denomination of the two systems is encountered from time to time.

σ−π Hyperconjugation p−d Homoconjugation

π pd Si C σ CC C Si

Scheme 27: Hyper- and Homoconjugation with Allylsilanes.

23 A number of studies have been made to shed light on the underlying mechanism of the β-stabilisation. Basically, the stabilisation can be explained by three factors: 1. induction effects from the Si-atom 2. hyperconjugation through vertical stabilisation 3. hyperconjugation through nonvertical stabilisation.

Induction can be seen as a lowering of energy from through-bond delocalisation of electrons and hyperconjugation as energy lowering by through-space delocalisation. The magnitude of induction should depend on the bond length but not the bond angle, while hyperconjugation should vary according to the dihedral angle between the vacant p- orbital and the C–Si bond. The inductive effect was from an early stage found not to influence the β-stabilisation markedly.86,89,90 Hyperconjugation, on the other hand, has been demonstrated to be the main source of the β-effect in ab initio studies.86 This was later corroborated in solvolysis experiments, where the direction dependence of the β- effect was beautifully illustrated with comparisons of molecules where the leaving group was placed either in a gauche 20 or antiperiplanar 21 position89,91 to the trimethylsilyl group (Scheme 28).

SiMe3 SiMe t-Bu 3 t-Bu

O(CO)CF3 O(CO)CF3 20 21 Gauche Conformation Antiperiplanar Conformation

Scheme 28: Gauche and Antiperiplanar Conformations in Solvolysis Experiments.

The conformer with the trimethylsilyl group in the antiperiplanar position 21 lost the leaving group 1012 times faster and the one with the gauche trimethylsilyl 20 group 3.3×104 times faster than the molecule without trimethylsilyl substitution, which amply demonstrates the importance of the dihedral angle for an effective β-stabilisation. The hyperconjugative factor for the stabilisation of the antiperiplanar compound was estimated to 1010 and the inductive factor to 102.89 The substituents of the silicon atom can also influence the β-effect. The presence of electronegative elements (e. g. –Cl) on the silicon atom leads to a lower degree of stabilisation.92

There has been a debate during the last fifteen years over the structure of the β-silyl allyl cation, where some proponents argue for an open, vertical structure88 22 and some for a bridged, nonvertical structure93,94 23 (Scheme 29).

24 R3 Si R3Si

C CR´2 H2CCR´2 H2 22 23

Open/Vertical Bridged/Nonvertical

Scheme 29: Vertical and Nonvertical Modes of β-Stabilisation.

If sufficient proof of the existence of the hyperconjugative effect has been easy to come by has the differentiation between vertical and nonvertical stabilisation been difficult to examine experimentally. A partially bridged system has been argued for in gas-phase experiments.93,94 Lambert has argued convincingly for the vertical mode of stabilisation in several papers88 where NMR-95 and dihedral angle studies90 were made in solution. The additional stabilisation of a second silyl group has been found to strengthen the β- effect as well.96 This would support the open structure as additive stabilisation should not be allowed in bridged structures.

The β-effect is also prominent in germanium and tin systems.92,95,97 The β-effects of germanium-stabilised allyl compounds are slightly higher than the corresponding silicon compounds, while the β-effect of tin is considerably higher than both. In fact, the rate- enhancement due to β-stabilisation of tin was so high that it could only be approximated with a weaker leaving group (OAc) and a gauche conformation in solvolysis studies.97

A couple of examples of where the β-stabilising effect of silicon has been applied synthetically include the generation of the relatively stable cation 24,98 which is 15 kcal/mol more stable than the corresponding dimethyl derivative and the higher stabilisation of the five-membered ring 25 compared to the six-membered 26,99 where the overlap of the lone pairs of the oxygen is more favourable in 25 than in 26. The rate 100 of protonation was also found to be higher in the alkyne 27 when X is –SiMe3 than when X is –H (kSiMe3/kH= 54,300) (Scheme 30).

Me3Si SiMe3 H H O = O = R3Si SiR R3Si R3Si 3 24 25 26

X H+ Me X Me H 27 Scheme 30: Reactions where β−Stabilisation has been Utilised.

25 The stabilisation due to hyperconjugation is not limited to the β-position and a few cases of γ-stabilisation83,101 have been reported as well.

2.3.2 The Chelation of Palladium to Allylamines and Allylamides. The palladium(II)–nitrogen chelation has been reported on many occasions; for example in conjunction with nucleophilic attacks on palladium-coordinated allylamines such as 28102-106 and Heck-couplings with vinyl ethers, such as in 29107 and 30108 (Scheme 31). L L Nu Pd NMe Li2PdCl4 2 NMe2 NMe2 Nu Nu O O 28 Ar Pd Ar Pd X N I N Me2 29 30

Scheme 31: Examples of Coordination between Palladium and Allylamines or Vinyl Amine Ethers.

Complexes such as 28 are reported to be quite stable and no decomposition was discovered at 100 °C in NMR studies.102 The palladium–nitrogen coordination can also be strong enough to allow the analysis of crystal structures such as 31109,110 (Scheme 32).

Me2 N Cl Pd Cl

MeO 31 Scheme 32: Crystal Structure of a Palladium-Homoallylamine Complex.

In complex 32 was the formation of 33, instead of the anticipated 34, explained by a coordination of the palladium moiety to the anilic, primary nitrogen, which hindered β- elimination and readdition and further migration of the double bond111 (Scheme 33).

HI Pd NH2 NH2 N N N Ar

MeO O MeO O MeO O 32 33 34 Scheme 33: Suppression of Double-Bond Migration through Palladium-Nitrogen Coordination.

26 The ring sizes of the chelated complexes are known to be important and a preference for five-membered rings over six-membered rings is well established.106 In sulphur- coordination experiments were complexes such as 36 spontaneously formed from six- membered intermediates 35 (Scheme 34).112

R SCH2C6F6 R SCH2CH2C6F6 Pd Pd HBr HBr

2 35 2 36 Scheme 34: Preferred Five-Membered Ring Size of a Palladium-Sulphur Complex.

Chelation of palladium to amines has also been investigated with IR spectroscopy,113,114 where the N–H stretching vibrations appeared at lower frequencies in palladium- coordinated than in free amines. Coordination has in some experiments been shown to increase in non-polar solvents.115

The electron density of the heteroatom in palladium–heteroatom coordination complexes may influence the strength of the chelation. In complex 37 was coordination stronger with a nitrogen substituent capable of electron donation, such as propane or methane, than with the electron withdrawing methyl ester (Scheme 35).116 In addition, the imine-substituted complex 38 underwent a transmetallation reaction via a Pd–N coordinated five-membered ring easier than the acetyl substituted 39, where resonance delocalisation in the amide group was reported to reduce the chelating ability of the nitrogen (Scheme 35).117

Pd Ar + ArX, Ag R N Pd(0) N Ar N R R 37 Ph R = COOMe, Alk Ph H CO C H3CO2C Ph 3 2 N Transmetallation NPh Pd -Bu SnBr Br 3 Pd Bu Sn 3 38 H CO C 3 2 NHAc

Pd Bu3Sn Br 39 Scheme 35: Influence of the Electron Density of Nitrogen on Palladium-Nitrogen Coordination.

27 Similarly was complex 36 in Scheme 34 more stable with n-Bu than with Ph as sulphur substituent (–R).112 Electron-withdrawing substituents in general facilitate ring opening in palladium–carbonyl complexes118 and an electron-poor palladium centre has been suggested to be more sensitive for coordination than electron-rich.115

Equilibrium studies with palladium chelated to the bidentate ligand bis(dimethyl- phosphino)ethane (dmpe) 40 (Scheme 36) revealed the following binding order for different degrees of nitrogen substitution NH2Et > NHEt2 > NH3 >> NEt3 and for different substituents NH2Et > NH2i-Pr > NH3 > 1-adamantamine > NH2t-Bu > 119 NH2Ph. The effect of the basicity of the amine ligand appeared to be negligible for amines with cone angles54,58 smaller than 120°.119 The same report found a stronger coordination when the nitrogen was enclosed in a ring system, which was explained by steric effects. Pyridine had an anomalously strong binding to palladium, where the planarity of the ring system was suggested to favour π-back bonding to the metal centre. All in all were both steric and electronic effects found to affect the binding of nitrogen to palladium.119

(H3C)2PP(CH3)2 Pd H3CNR2 40 Scheme 36: Equilibrium Study of Nitrogen Ligands to Palladium.

In the case of allylamides can the situation be more complex, as chelation is possible both to the nitrogen and the carbonyl group. Studies on small peptides indicate that when a transition metal coordinates to the nitrogen of a secondary amide can a deprotonation of the nitrogen occur more readily, which would generate an anionic nitrogen, possibly more capable of coordination to Pd2+ than a neutral nitrogen.120 The ease of deprotonation is dependent on the metal and representative pKa values for a number of transition metals have been reported (Pd2+=2; Cu2+=4; Ni2+=8; Co2+=10).120 If the metal–N coordination can generate a five-membered ring is deprotonation easier than with six-membered intermediates. Furthermore, the charge of the amide is important for the mode of coordination, as anionic, deprotonated forms of amides in small peptides coordinate at the nitrogen and neutral forms at the carbonyl.120

28 Palladium–carbonyl coordination has been suggested on many occasions.106,118,121-123 Hegedus has suggested structure 41 as an intermediate in nucleophilic attacks on vinyl carbamates121 and Alvisi and Blart presented the six-membered Pd–carbonyl coordinated structure 42 and the five-membered Pd–N coordinated 43 as alternative intermediates in the arylation of Boc-protected allylamines with a vinylic trimethylsilyl group (Scheme 37).122

L Ot-Bu L Ot-Bu Pd L - Nu L Nu Pd O NH O HN L L L L O HN Ar O Pd Pd NH I OBn I OBn Ar SiMe3 41 SiMe3 42 HBoc N H Boc L L L L N Pd Ar Pd I I Ar SiMe Me3Si 3 43

Scheme 37: Examples of Palladium Coordination to Carbonyls and Nitrogen.

Impressive reports of rhodium–carbonyl coordination also support the importance of five-membered structures.124

29 3 Recent Developments

3.1 Microwave Chemistry

3.1.1 The Microwave Oven The first examples of the use of microwaves to enhance reaction rates in organic chemistry date as far back as when the first household microwave ovens were introduced, but the difficulty of controlling the reactions lead to unreproducible reactions and explosions. Domestic ovens have been developed to meet different standards than what is needed in chemistry and use rotating dishes or other mechanisms to spread the microwaves in the oven to reduce the risk of unevenly heated food rather than trying to focus the radiation. When doing chemistry in multi-mode ovens there is thus a risk of getting poorly reproducible results as the heating is not uniform and this, in turn, accounts for the risk of explosions.

Several factors can account for the momentum gained in the development of microwave chemistry during the last few years. One reason is the advent of combinatorial chemistry and high-throughput screening, where there is a need for high-speed synthetic methods. Another is the development of single-mode microwave ovens, where not only the power input can be controlled but also, what is perhaps more important, the focusing of the microwave irradiation. Single-mode reactors found one of their first applications in the synthesis of radiopharmaceuticals, where a fast, controlled synthesis of radioactive compounds were of paramount importance.125,126

The possibility of controlling the power input in single-mode ovens is a great improvement for the synthetic chemist, as even though it appears as if the temperature can be controlled in the domestic oven, it is mostly not the power input that is changed but only the length of the heating cycles. Even so, in spite of the differences between the focused single-mode and domestic multi-mode ovens have the latter been used in most of the papers published as of today. The danger of explosion is minimised by several methods, some of which are running the reactions without solvents in open vessels, although the use of solvent-free conditions is reported with single-mode reactors as well,127 and modifying the domestic oven with pressure valves. The popularity of domestic microwave ovens is an effect of the considerably higher prices for the single- mode apparatus combined with the availability of the multi-mode ovens. In fact, there are not many companies that manufacture single-mode ovens commercially, but the need for controlled and fast synthetic processes has lately created a substantially higher demand for single-mode ovens in combinatorial chemistry.128 There is also a possibility that microwave chemistry can be used to reduce the environmental damage that often is associated with chemistry at an industrial scale.129 The development of single-mode

30 microwave chemistry is still dynamic and it is at present difficult to predict its precise development in the future.

3.1.2 General Physics Electromagnetic radiation, such as ultraviolet light, is well known to be able to transfer energy. Microwaves, which is a form of electromagnetic radiation with a frequency range of 0.3–300 GHz (Scheme 38),130 is no exception and can be used to transfer energy as long as the medium that is to be heated is able to absorb radiation in the microwave range.

Scheme 38: The Electromagnetic Spectrum.

In order to avoid disturbance with telecommunications and radar, which both utilise frequencies in the microwave range, have two frequencies been allotted for microwave oven use: 900 MHz (λ = 33.3 cm) and 2.45 GHz (λ = 12.2 cm). Most commercial ovens, both single- as well as multi-mode, use the 2.45 GHz frequency.

The term “dielectric heating” is frequently used when microwave heating is discussed. A dielectric material contains either permanent or induced dipoles, which can function as a capacitor that can store charge when placed between two electrodes.131 Heating can be induced, mainly through polarisation and ion conductance, when microwave irradiation is applied. The polarisation consists of several parameters:130

31 αt = αe + αa + αd + αi where the total polarisation (αt) is described by the sum of the electronic (αe), atomic (αa), dipolar (αd) and interfacial (αi) polarisations. These different forms of polarisation have different speeds with which they can be realigned in an oscillating microwave field. The electronic and atomic polarisations are not of interest in microwave chemistry as the response time, or lag factor τ, due to αe and αa is considerably faster than the frequency of the field and as a consequence is no, or very little, energy transferred to the medium as a result of these effects. The lag time for the dipolar polarisation (αd) is, on the other hand, similar to the microwave frequency. The polarisation can then lag behind the oscillation of the field and electromagnetic radiation can be converted into heat.130,131

The efficacy with which a certain medium can be heated by means of microwave irradiation can be described by the factor tan δ, the dielectric loss tangent, which is defined as the quota of the dielectric loss ε’’ (the efficiency with which electromagnetic radiation can be converted into heat) and the dielectric constant ε’ (the ability of a molecule to be polarised by an electric field).

tan δ = ε’’/ε’

The tan δ value can thus describe the ability of a material to convert electromagnetic energy into heat at a given frequency and temperature.130 It should be noted that the tan δ value could in some liquids vary markedly with temperature. In the examples of water and ethanol is the tan δ value considerably smaller at higher temperatures. A list of the tan δ values of some common solvents has been published by Mingos.131 The choice of solvent can, as a consequence, be of great importance when the reaction conditions for microwave-heated reactions are considered and nonpolar solvents, with low tan δ values, should in most cases be avoided.

The other mechanism through which heating can be transferred to a reaction medium is ionic conductance. The addition of salts to increase the ionic conductance or a minor amount of polar solvent can in many cases suffice to raise the ability of the reaction medium to absorb microwave irradiation if unpolar solvents are vital for the reaction.131 When an oscillating microwave field is irradiated through a solvent with ions will the ions also oscillate and transfer heat to the reaction medium through friction. It should for safety reasons be pointed out that the heating due to ionic conductance has a tendency to increase with higher temperatures and if care is not taken can the temperature rise unpredictably fast (the so called “thermal runaway effect”) with a concomitant risk of reaction vessel rupture and explosion.

32 Interfacial polarisation (αi) can occur as a build-up of charges takes place between liquid-liquid interfaces; the so-called Maxwell-Wagner effect. This effect is not as well understood as the dipolar polarisation but in an examination of interfaces between 132 soybean oil and water has αi been suggested to be important. Microwave irradiation has also proved to be of use in phase-transfer catalysis,133 where the interface area should be large.

3.1.3 Rate Enhancements with Microwave Chemistry The sometimes remarkable rate enhancements reported with microwave-heated reactions have fuelled a debate whether there is a specific effect on the reacting molecules from the microwaves, apart from the increase in temperature, that speeds the reaction (the so called non-thermal microwave effect). This effect could be explained, for example, by storage of microwave energy in the vibrational energy of a molecule (enthalpic effect) or by an alignment of molecules (entropic effect).134 An increasing amount of reports135,136 seem to disfavour the non-thermal microwave effect as the energy transmitted by the microwaves is too weak to cleave a chemical bond (approximately 1 J/mol,135 0.3 kcal/mol137). Stuerga and Gaillard have presented a number of conclusions for why the non-thermal microwave effect is non-existent.135 The three most important of which are: 1. The energy of the microwave photon is too low to break chemical bonds (10-5 eV135). 2. The electric field strength in microwave ovens cannot induce shifting in chemical equilibria. 3. The electric field strength is too low to induce organisation. Dipolar moments remain randomly distributed due to thermal convection.

If we are to assume that the non-thermal microwave effect is imaginary there must be other important factors that can account for the rate-enhancements of microwave chemistry (thermal microwave effects). Mingos has presented four advantages of microwave heating, compared to conventional, thermal heating, which are discussed below:131 1. A very high heating rate, compared to thermal reactions, can be achieved if the microwaves can interact with at least one component in the reaction mixture. 2. The nature of the electromagnetic radiation allows for heating where there is no contact between the heating medium and the heated material. For the synthetic chemist this phenomenon is most importantly embodied in the absence of wall- effects with microwave heating. The wall-effect is created when a reaction vessel of, in most cases, low temperature is put in a reaction medium, such as an oil-bath, with

33 high temperature. A temperature gradient will be created in the reaction mixture, where the temperatures closest to the walls of the vessel may be considerably higher than the temperatures in the bulk of the reaction mixture.129 In palladium chemistry,138 this has been suggested to be a source of catalyst deactivation and the prospect of doing chemistry without wall-effects is thus promising for applications in organometallic chemistry. 3. Selective heating can be attained when different materials or reactants interact to a different degree with microwave irradiation. In the specific case of organometallic- and palladium chemistry has the possibility of hot spots136 around the metal catalyst136 been discussed. The local higher temperatures around the catalytically active species should, if the complex can withstand the higher temperatures, result in faster reactions. 4. The selective interactions mentioned above also imply that microwave chemistry is well suited for reactions conducted under high pressure. “The pressure cooker effect” is frequently hailed as one of the main factors behind the rate enhancement of reactions performed in closed vessels. A dramatic increase in turnover numbers has been reported in palladium-catalysed reactions conducted under high-pressure conditions.139 Running a reaction in a closed vessel under increased pressure presents the opportunity of heating the reaction medium above the normal boiling point of the solvent. Depending on the solvent used can the boiling point be raised significantly over the boiling point at atmospheric pressure140,141 and the higher temperatures in themselves are capable of inducing faster reactions. Gedye and Wei have investigated the rate enhancement of microwave irradiation at atmospheric pressure142 and draw the conclusion that rate-enhancements of synthetic reactions do not occur as easily under microwave heating in open vessels and that no non- thermal microwave effect can be discerned. The same authors refer the slight rate enhancements that at some instances were found to the presence of local hot spots in the reaction medium.

34 3.2 Fluorous Chemistry

3.2.1 History The characteristic of fluorine that is taken advantage of in fluorous chemistry is the inability of perfluorinated or highly fluorinated compounds to be dissolved in organic solvents or water.143 A separation in an ordinary separation funnel with a perfluorinated solvent, such as FC-84, dichloromethane and water gives rise to a three-phase system with the fluorinated phase at the bottom. The physical bases for the poor solubility of fluorinated solvents are their low surface tensions, low intermolecular interactions, high densities and low dielectric constants. They are also nontoxic, stable (due to the stability of the C–F bond) and, unlike the freons, relatively environment-friendly.144 Horváth and Rabái introduced the term “Fluorous Biphasic Systems” (FBS) where fluorinated rhodium catalysts were used in the hydroformylations of olefins.145,146 The concept was further developed by Curran, who presented synthetic strategies where fluorinated tags were used to introduce samples in the fluorous phase with the aim of ameliorating the purification step (Scheme 39).147-149

Reagents F + Substrate F Substrate F Product + excess Reagents

Purification F Product Detachment F + Product

Scheme 39: Example of Curran's Fluorous Synthetic Strategy.

One further benefit of this strategy is the possibility of recycling expensive ligands and likewise expensive metal catalysts bound to the ligands.150,151 Once the tag or ligand is isolated in the fluorous phase can the fluorous solvent, which has a low boiling point due to its low polarisability and low intermolecular forces, easily be removed under reduced pressure and collected. Fluorinated solvents have other interesting properties as well, such as a high ability to dissolve oxygen gas. This has been taken advantage of in medicinal technology, where fluorinated solvents have been used as oxygen carriers in the treatment of lung failure.152

3.2.2 Applications in Modern Chemistry Due to the insolubility of highly fluorinated compounds in water or organic solvents have various synthetic strategies been used in order to allow the full mixing, a prerequisite for successful chemistry, of the fluorous with the non-fluorous reagents. Fully fluorinated compounds are thus for solubility reasons seldom used but rather tags with a varying degree of fluorination.149,153-155 Large compounds with a high molecular

35 weight need a higher fluorine content in the attached tag in order to be fully distributed to the fluorous phase than small compounds with low molecular weights.156,157 Indeed, the mass percentage of fluorine atoms compared to the total amount of non-fluorous atoms can usually give a good approximation of the solubility in fluorous media. Horváth has recommended a fluorine content of at least 60% in order to ensure a good partitioning to the fluorous phase.146 A recent addition to the discussion is a report by Gladysz where the fluorous phase affinities for a number of benzenoid compounds with varying degrees of fluorination are investigated.158

Different methods have also been used in the measurement of how fluorous a compound is. Rabái introduced the term fluorophilicity (f),159 where f = lnP and P is the partition coefficient of the compound over perfluoro(methylcyclohexane) and toluene at 25 °C while Horváth and Hughes have analysed fluorous partition coefficients PFBS = cFluorous phase/cOther phase, where the perfluorinated heptane was used as the fluorous phase and toluene as the organic.160

It has been found that, when fluorous tags are used, the inductive power of the fluorous atoms can influence the reactivity of the tagged molecule. In order to prevent this is often a spacer introduced at the end of the tag, where it connects with the reactant or product (Scheme 40).146 Spacer Fluorous Tag

R-Sn(CH2CH2C10F21)3

Scheme 40: Stille Reactant with Three Fluorous Tags.

Fluorinated alkyl tails separated from the rest of the tag with a spacer have been proven to be more advantageous than fluorinated aryl groups, partly due to the risk of aromatic interactions when the latter is used.161 The ideal length of the spacers has been the subject of several papers.153,154,162 Some reports question if the ethylene spacer is long enough to insulate properly154 and other have found the propylene spacer more unstable.162 The use of longer spacers does not seem to be of value and may increase the risk of micelle creation.146 The inductive effect of the perfluorinated tail has also been used in reactions where fluorous tags were found to react faster than non-fluorous tags,163,164 although there are other reports where the reactivity is similar162 or lower165 with fluorous compounds.

36 Another parameter that can be adjusted to improve the solubility and mixing of fluorinated and non-fluorinated molecules is the solvent. Hybrid solvents, such as BTF 166 (C6H5CF3) with fluorous as well as organic properties, can dissolve organic as well as most fluorinated molecules, especially at elevated temperatures. One effect that is regularly noted in fluorous chemistry is the insolubility of the fluorous phase with the hybrid or non-fluorous phases at room temperature. When heating is applied will the solubility of the two phases increase and a one-phase system appears that upon cooling to room temperature will recreate the two-phase system. If the solubility of the fluorinated phase is too poor, most often due to a high degree of fluorination, may the intermixing of the phases not occur, with an ensuing synthetic failure.

It is from this discussion possible to reach the conclusion that in order for a successful fluorous synthetic strategy to be realised must the fluorine content of the tags be controlled; too high a degree of fluorination may result in an insoluble fluorous phase that cannot react with non-fluorinated reagents and a too low degree will, on the other hand, give leaching of the fluorinated content into the organic phase. An important and interesting application of lightly fluorinated tags has been applied by Curran in the use of fluorous reverse phase (FRP) chromatography.167-169 In this context are molecules with a low degree of fluorination beneficial, as a good separation can be fulfilled by eluting non-fluorous compounds with 80% MeOH/H2O and later the fluorous compounds with acetonitrile.169

If the reaction conditions and choice of reagents can be controlled to ensure a smooth separation can fluorous chemistry be used to good effect in synthetic strategy.149 When toxic compounds, for example the tin reagents used in the palladium-catalysed Stille– reactions, have to be used, it would be not only convenient but also environmentally advantageous if the toxic compounds could be easily separated from the reaction mixture. Curran has used fluorous tags on tin reagents that enable not only the separation of the tin compounds but also the possibility of recycling them for further synthesis.151 This is of importance as tin compounds are difficult to isolate from the reaction mixture.170 In radical-mediated reactions has the use of catalytic amounts (10 171 %) of tin hydride in combination with the reducing agent NaBH3CN been shown to give full conversion in a coupling of adamantyl bromide with acrylonitrile.151

37 4 Aims of the Present Study

The aims of the present study were: a) To study and improve the regiocontrol in the Heck-reaction by developing novel synthetic methods of arylation at the internal carbon of selected allylic systems of particular interest for medicinal chemistry. The allyltrimethylsilane compounds were chosen due to their synthetic usefulness and the allylamine substrates for their potential as bioactive compounds. Furthermore, the heteroatoms of the two types of olefins, encompassing C=C–C–Si and C=C–C–N fragments respectively, were foreseen to exert an impact on the regiocontrol of the Heck-reaction that was thought to be important to assess. b) To use the internal arylation of allylic substrates to develop short and attractive synthetic strategies for MAO-B and SSAO inhibitors. c) To investigate the effects of single-mode microwave heating on chemical reactions with the aim of reducing the reaction times without compromising the regioselectivity and efficacy of the reactions.

d) To explore the use of heavily fluorinated tags and their applications in microwave- heated reactions.

38 5 Results and Discussion

5.1 Highly Regioselective Internal Arylation of Allyltrimethylsilane (Paper I)

β-Arylated compounds of the type 45 had previously been synthesised using monodentate ligands and aryl iodides, e. g. by regioselective phenylation of 44 at 120 °C in the presence of silver salts (Scheme 41).46

I SiMe3 Pd(OAc)2, PPh3 + Me3Si SiMe3 AgNO 120 oC 44 3, 45

SiMe3 SiMe3

46 47

Scheme 41: Previous Syntheses of Arylallyltrimethylsilanes.

Allyltrimethylsilane was arylated terminally, producing mainly 46, by palladium and monodentate ligands with aryl halides as aryl precursors. The addition of silver salts, which was thought to produce a cationic palladium by halide scavenging (Scheme 13), produced the double-bond isomer 47 and small amounts of 45 (Scheme 41).46 A study was undertaken to investigate if the direct synthesis of compounds 45 could be brought about through a regioselective Heck-coupling of aryl triflates with allyltrimethylsilane and a palladium/bidentate ligand catalytic system (Scheme 42).

OTf SiMe3 SiMe Pd(OAc)2 SiMe3 + 3 + dppf R R R 48 45 46

Scheme 42: Palladium/dppf-Catalysed Internal Arylation of Allyltrimethylsilane.

Several bidentate ligands were investigated (dppp, BINAP, dppb, dppe) but none gave the same high and reproducible regioselectivities as dppf.55,56 Monodentate ligands invariably produced very poor regioselectivities. The reaction can due to the choice of aryl triflates as aryl precursors and a bidentate ligand be taken to go through the cationic pathway. The polarisation of the double bond of the allyltrimethylsilane would lead to a preferred arylation at the internal carbon of the olefin, partly due to the greater ability of the internal, secondary carbon compared to the primary, terminal carbon to stabilise a

39 burgeoning positive charge and partly due to the documented β-stabilising effect of the silicon group (Scheme 43). The transition state 49 in Scheme 43 is depicted to underline the β-stabilising effect and is not a true estimation of the insertion process, which is believed to be concerted.39,40

L L SiMe3 L L L L L L Pd Pd Pd Pd SiMe3 Ar Ar X -X Ar Ar γ β Ar X = OTf, OAc α Me3Si Me3Si Me3Si 49 Scheme 43: Internal Arylation of Allyltrimethylsilane.

Aryl triflates with a wide range of functionalities could be coupled with high regioselectivity and moderate to high isolated yield (Scheme 44). The electron-poor aryl triflates 48g and 48h and the ortho-substituted 48c needed a higher reaction temperature (80 °C vs. 60 °C for the other aryl triflates) to reach full conversion in over-night reactions. Most couplings produced a small amount of the terminally, γ-arylated 46, in addition to the desired 45 (Scheme 42).

The yields were lowered due to formation of side-products by a) aryl migration, b) deoxygenation of the aryl triflates and c) desilylation of the products.

Aryl migration occurred mainly with the electron-rich aryl triflates (48a and 48b). It was also noted in line with earlier results67 that the ortho-substituted electron-rich aryl triflate 48c gave a lower degree of aryl migration compared to the para-substituted electron-rich 48a and 48b.

Deoxygenation of the aryl triflates occurred particularly with the electron-poor aryl triflates. This side reaction could be reduced by switching from an organic base (Et3N) 75 to an inorganic (K2CO3) as one source of hydrides for the deoxygenation of triflates could be amine bases (Scheme 24).77 The triethylammonium formate-catalysed deoxygenation of triflates has been found to occur more easily when electron-poor aryl triflates and non-coordinating solvents are used.172 Aryl triflates are also more readily reduced by electrochemical reactions when substituted by electron-withdrawing groups.173

Desilylation of the product to the corresponding α-methyl-styrene derivatives was noted especially in the reactions conducted at 80 °C (entries 3, 7–8; Scheme 44).

40 Thermal Reactions Microwave-Heated Reactions

Reaction a Isolated a GC-MS Entry Product Selectivity Temp. Yieldb Time/ Selectivity Yield 45/46 (oC) Microwave 45/46 45+46 Powerc 45+46

e 1. SiMe3 97/3 60 42% 5 min/50 W 78/22 34% 45a MeO

f 2. SiMe3 94/6 60 47% 5 min/50 W 88/12 69% 45b t-Bu

d e 3. Me SiMe3 100/0 80 60% 5 min/50 W 93/7 66% 45c Me Me 4. SiMe3 95/5 60 67% 5 min/50 W 92/8 62% 45d

SiMe3 5. 98/2 60 77% 5 min/50 W 97/3 66% 45e

6. SiMe3 98/2 60 85% 7 min/50 W 97/3 54%

Cl Cl 45f

d 7. SiMe3 94/6 80 31% 7 min/60 W100/0 48% O 45g

Me d d 8. SiMe3 100/0 80 59% 10 min/50 W 100/0 28% 45h NC

The thermal reactions were run in 2.5 mmol scale under nitrogen atmosphere at 60 oC (entries 1-2, 4-6) or 80 °C (entries 3, 7-8) with 1 equiv of 48a-h; 0.03 equiv of Pd(OAc)2, 0.132 equiv of dppf, 5 equiv of allyltrimethylsilane; 2 equiv of base (Et3N for entries 1-5, K2CO3 for entries 6-8) and 10 mL of fresh acetonitrile. All reactions were completed after 20 h. The microwave reactions were run in 1.0 mmol scale in sealed Pyrex tubes with septa under nitrogen atmosphere and heated by means of microwave irradiation. The microwave heated reactions were run in 1.5 mL of fresh acetonitrile with 1 equiv of 48a-h, 2.5 equiv of allyltrimethylsilane, 0.03 equiv of Pd(OAc)2, 0.132 equiv of dppf and 2 equiv of base (Et3N for a entries 1-5, K2CO3 for entries 6-8) Selectivity internal β-arylation/terminal γ-arylation. Determined by GC-MS and 1H-NMR. b>95% by GC-MS. cContinuous irradiation at 2450 MHz. dOnly product 45 detected by GC-MS. eThree isomeric products detected by GC-MS. fFour isomeric products detected by GC-MS.

Scheme 44: Internal Arylation of Allyltrimethylsilane with Aryl Triflates 48.

41 The regioselectivities of the arylation of allyltrimethylsilane were found to be high and a comparison with 4-methyl-1-pentene, an olefin where the β-effect of silicon does not influence the regioselectivity, indicated a higher tendency in the case of allyltrimethylsilane to favour arylation at the internal β-carbon (Scheme 45). The arylation of 4- metyl-1-pentene generated four products with the right m/z as analysed by GC–MS and the selectivity for internal arylation was gauged by hydrogenation of the double bonds with Pd/C and H2 and subsequent analysis of the two resulting internally and terminally arylated isomers by GC–MS and 1H–NMR.

95 86

SiMe 5 3 14

Scheme 45: Comparative Regioselectivities in the Phenylation of Allyltrimethylsilane and 4-Methyl-1-Pentene.

In a competitive experiment where one equivalent each of allyltrimethylsilane and 4- methyl-1-pentene were added to the reaction mixture was the allyltrimethylsilane arylated 5.5 times more easily than the alkene, which might indicate a reduced energy barrier, presumably due to β-stabilisation, in the rate-determining insertion step.

A series of single-mode microwave heated reactions was also performed (Scheme 44). The reaction times could be effectively reduced from thermally heated over-night reactions (<20 h) to 5–10 minutes with microwave heating; a truly dramatic reduction in reaction time without far-reaching consequences for the outcome of the reaction. The regioselectivities under microwave heating were generally slightly lower than the corresponding thermal reactions, except for the synthesis of 45g, where the regioselectivity was raised from 94/6 (thermal heating) to 100/0 (microwave heating). The electron-poor aryl triflates needed longer reaction times and higher microwave input than the electron-rich aryl triflates, which is reminiscent of the tendency in the thermal reactions, where the electron-poor aryl triflates also needed higher temperatures.

The lower regioselectivity noted in many of the microwave-heated reactions could be caused by the problem of controlling the reaction temperature. Under microwave- irradiation can the reaction temperature be raised high above the boiling point of the solvent and in many cases were visible signs of catalyst decomposition present in the reaction mixtures (palladium-black, palladium-mirror). It could be seen in thermal reactions, where the temperature could be controlled more easily, that when the reaction temperature was raised from 60 °C to 80 °C was also the regioselectivity in many cases better. There are thus reasons to assume that there is a delicate balance between a faster

42 and more efficient Heck-coupling with a higher regioselectivity (by a moderately higher temperature) and catalyst decomposition (due to too high temperatures) and, presumably in the case of microwave-heated reactions, hot-spots around the catalyst metal. This balance was difficult to control with the present microwave cavity, where the reaction temperature could not be controlled by any other means than by the microwave power input and reaction time.

To further evaluate the influence of the β-effect on the regioselectivity in the Heck- reaction were couplings made where the nature of the olefin and the triflate were subject to variation (Scheme 46).

Selectivity Selectivity Product Product (Internal/Terminal) (Internal/Terminal)

SiMe SiMe 3 3 71/29 NR 50 52

SiMe3 89/11 75/25 51 53

Scheme 46: Regioselectivity in the Heck Reaction.

Couplings with phenyl triflate and a Pd(OAc)2/dppf system with vinyl silane and 3,3- dimethyl-1-butene resulted in the formation of 50 and 51 in poor regioselectivities; the synthesis of 50 also produced a considerable amount of styrene. The arylation of 50 could be expected to result in a mixture of regioisomers as the polarisation of the olefin should favour internal- and the β-stabilisation terminal arylation. The 3,3-dimethyl-1- butene induced a somewhat higher ratio of internal arylation. The terminally disubstituted 1,1-dimethyl-3-(trimethylsilyl)-1-propene was unreactive and no trace of 52 could be detected by GC–MS even at elevated reaction temperatures. The use of vinyl triflates, such as the cyclohexenyl triflate, produced 53 along with a mixture of products (Scheme 46) and changing the bidentate ligand dppf to four equivalents of the monodentate ligand PPh3 did not result in a regioselective synthesis of 46d but in a complex mixture of stereo- and regioisomers. Couplings under our standard reaction conditions with aryl bromides in the presence of thallium- or silver salts43 instead of aryl triflates also resulted only in complex mixtures of stereo- and regioisomers. If two equivalents of bidentate ligand was used instead of 4.4 equivalents were the reactions considerably slower (2–3 days) and the yields tended to be lower although the regioselectivities were largely unchanged.

43 A highly regioselective method for the synthesis of the terminally arylated 46 under phase-transfer conditions at room temperature was recently published, where small changes in the reaction conditions could allow the formation of the desilylated product 54 in high yield with no detectable production of 46 (Scheme 47).174

Pd(dba)2, DABCO SiMe3 n-Bu NCl I 4 46 + SiMe3

Pd(dba)2, cat. PPh3 54 n-Bu4NOAc

Scheme 47: Synthesis of 46 and 54 by Heck-Arylation under Phase-Transfer Conditions.

We made a considerable effort to synthesise the internally arylated allyltrimethylstannane 55 (Scheme 48) from aryl triflates and allyltrimethylstannane as the β- stabilisation of tin is reported to be considerably higher than that of silicon.97

OTf SnR3 Pd(OAc)2, dppf SnR3 + K2CO3, CH3CN 55

Scheme 48: Failed Internal Arylation of Allyltrimethylstannane.

The present reaction conditions were evidently not suitable for the synthesis of internally arylated allylstannanes as no product formation was seen by GC–MS analyses. A high degree of destannylation and formation of α-methylstyrene was noted. The product 55 could easily be synthesised by lithiation175 as in Scheme 49, proving that the instability of 55 was not the only reason for the failure of the palladium-coupled reaction.

1. BuLi, KOt-Bu SnMe3

2. ClSnMe3 55

Scheme 49: Synthesis of Internally Arylated Allyltrimethylstannanes by Lithiation.

In short, the bidentate-controlled arylation of allyltrimethylsilane with aryl triflates was shown to produce the internally arylated 45 with good regioselectivity. This reaction complements our previously published syntheses of the isomers 46 and 47, using monodentate ligands.

44 5.2 Highly Regioselective Internal Arylation of N,N-Dialkylallylamines and Primary Allylamine Equivalents (Paper III and IV)

The success of the internal arylation of allyltrimethylsilanes encouraged the investigation of internal arylation of other allylic substrates. In our medicinal chemistry program we had had the need to insert amine functionalities with different chain lengths from an aryl backbone and the synthesis of compounds 56 (Scheme 50) appeared to be of high interest as primary allylamines had been reported to yield chiefly polymers in Heck-couplings.176 Pd(OAc)2, dppf, Base, CH CN or OTf 3 NR2 DMF + NR2 80 oC, 20 h or R microwave 3-5 min R 48 R=Alkyl 56

Scheme 50: Internal Arylation of N,N-Dialkylallylamines.

It was soon apparent that the regioselectivities in the arylations of N,N-dialkylallylamines were higher than the regioselectivities encountered with allyltrimethylsilanes and a different reaction mechanism was anticipated. The internal arylation of N,N-dialkylallylamines could go through an energetically favoured five-membered ring system 57, generating compounds 56 in excellent regioselectivity (Scheme 51).

L LL NR2 L L L Pd Pd Pd NR2 Ar Ar Ar X -X NR NR2 2 Ar X=OTf, OAc, Solvent 57 56 R= Alkyl

Scheme 51: Coordination-Controlled Internal Insertion of N,N-Dialkylallylamines.

A number of different aryl triflates were coupled with N,N-dimethylallylamine with very high regioselectivities and moderate to high isolated yields under thermal as well as single-mode microwave heating; the latter method proved efficient, just as in the reactions with allyltrimethylsilane, to reduce the reaction times to a couple of minutes (Scheme 52). The deoxygenation of the aryl triflates that occurred under the arylation of allyltrimethylsilane hampered the reactions with N,N-dimethylallylamine as well. The use of 1,2,2,6,6-pentamethylpiperidine (PMP) as base resulted in higher isolated yields in a few cases, but when deoxygenation occurred was the inorganic potassium carbonate superior (Scheme 52). When a change of base did not increase the yield was a higher amount of catalyst effective, although this, in turn, resulted in a higher degree of aryl migration, especially with the electron-rich aryl triflates.

45 Thermal Heatinga Microwave Heatingb

c Isolated Time (min)/ Isolated Entry Product Base β/γ Method β/γc Yield 56 (%)d Effect (W) Yield 56 (%)d

MeO NMe2 1. K2CO3 92/8 A 41 3/20 92/8 35 MeO 56i

NMe2 2. K2CO3 94/6 A 40 5/20 95/5 30 MeO 56a

NMe2 e 3. K2CO3 99/1 A 49 5/20 100/0 46 OMe 56j

NMe 4. 2 K2CO3 93/7 B 35 5/15 84/16 21 56b t-Bu

Me NMe2 e 5. Et3N 100/0 C 81 5/15 96/4 48 56c Me Me

NMe2 e 6. Et3N 100/0 C 42 5/15 88/12 24 56d

NMe2 e e 7. K2CO3 100/0 B 71 5/15 100/0 42 56e

NMe 8. 2 PMP 100/0e D 38 f Cl Cl 56f

NMe e f 9. 2 K2CO3 100/0 B 48 O 56g Me

10. NMe2 PMP 100/0e D 50 f 56h NC

aThe thermal reactions were run in 2.5 mmol scale under nitrogen atmosphere at 80 oC with 1 equiv of 48a-j; Pd(OAc)2, see Methods below; dppf, see Methods below; 5 equiv of N,N-dimethylallylamine; base, see Methods below; and 10 mL of acetonitrile. Method A (entries 1-3): 0.03 equiv of Pd(OAc)2, 0.132 mmol of dppf, 1.5 equiv of K2CO3. Method B (entries 4, 7, 9): 0.06 equiv of Pd(OAc)2, 0.264 equiv of dppf, 1.5 equiv of K2CO3. Method C (entries 5-6): As Method A but with 2 equiv of triethylamine instead of K2CO3. Method D (entries 8, 10); As Method b A but with 2 equiv of PMP instead of K2CO3. All reactions were completed after 20 h. Continuous irradiation at 2450 MHz. The microwave heated reactions were run in 1.0 mmol scale in septum-sealed Pyrex tubes with 1 equiv of 48a-j, 3 equiv of N,N-dimethylallylamine, 0.03 equiv of Pd(OAc)2, 0.132 equiv of dppf and 1.2 equiv of c 1 K2CO3 in 1.0 mL of DMF. Selectivity internal β−arylation/terminal γ-arylation. Determined by GC-MS and H-NMR. d>95% by GC-MS. eOnly one product detected by GC-MS. fNot isolated.

Scheme 52: Internal Arylation of N,N-Dimethylallylamine with Aryl Triflates 48.

46 The allylamines 3-piperidino-1-propene and 3-morpholino-1-propene were phenylated with excellent (>99.5/0.5) and moderate (88/12) regioselectivity, respectively, with phenyl triflate giving the products 58 and 59 in 50% and 56% isolated yield (Scheme 53). The lower regioselectivity of 59 might be the result of a competing coordination of palladium to the oxygen in the morpholino moiety.

O N N 58 59 β/γ >99.5/0.5 β/γ = 86/14 Isolated Yield: 50% Isolated Yield: 56%

Scheme 53: Internal Phenylation of 3-Piperidino-1-Propene and 3-Morpholino-1-Propene.

The internal arylation of N,N-dialkylallylamines was found to be sensitive of high temperatures and the raising of the reaction temperature above 80 °C invariably resulted in poorer yields due to formation of palladium-black and tar. Polymers have previously been noted as a side-product in palladium-catalysed reactions with allylamine.176

The microwave-heated reactions with the electron-poor aryl triflates were difficult to optimise compared to the thermally heated couplings due to the poor control of the reaction temperature. Temperature curves were made with fluoroptic probes in reactions run under microwave irradiation with DMF as solvent, indicating a reaction temperature close to 200 °C at 15 W, microwave power (Scheme 54).

250

200

150

100

50 Degrees Centigrade

0 0 60 120 180 240 300 360 420 480 Seconds

Scheme 54: Temperature Curve for Entry 5, Scheme 52.

47 A tendency for the thermally heated internal arylations of N,N-dimethylallylamines to yield higher regioselectivities with sterically unhindered electron-poor aryl triflates (entries 9–10, Scheme 52) than with sterically unhindered electron-rich aryl triflates (entries 1–2, 4; Scheme 52) could tentatively be explained by an inductive effect from the aryl group (Scheme 55). Partially sterically hindered aryl triflates (entries 3, 5, 7–8; Scheme 52) generally give higher regioselectivities than unhindered aryl triflates. Substrates with a high degree of steric hindrance (e. g. 2,6-dimethylphenyl triflate) give no product.

LL LL 2+ 2+ Pd NMe2 Pd NMe2

60 61 EWG EDG

Scheme 55: The Effect of the Electron Density of the Aryl Triflate on the Regioselectivity.

The higher regioselectivities for couplings with electron-poor aryl triflates are difficult to explain. Aryl groups with electron-withdrawing substituents 60 might through induction produce a slightly more electropositive palladium that would rather add to the terminal carbon as the internal position of the olefin easier could support a lower charge density than the terminal. Aryl groups with electron-donating substituents 61 would create the opposite effect. Alternatively could the electropositive palladium in 60 coordinate tighter to the nitrogen and create a more stable intermediate than 61 and thus induce a slightly higher regioselectivity. This inclination of electron-poor aryl triflates to give a higher regioselectivity is not as clear in the microwave-heated reactions due to the difficulty of controlling the reaction temperature under microwave irradiation. The internal arylations of allyltrimethylsilanes do not show any marked difference in regioselectivities between electron-rich and electron-poor aryl triflates.

Compounds with the general structure 62 have in the search for new anti-tumour agents received attention as protein tyrosine kinase inhibitors177 and epidermal growth factor- receptor inhibitors178 (Scheme 56). O

NR2

R 62 Scheme 56: General Structure of a Group of Anti-Tumour Agents.

48 Although carbonylative Heck-couplings with allylamines to the best of our knowledge have not been reported, we nevertheless decided to test the reaction as it, if it were to succeed, would be a very convenient synthesis of 62. Instead of 62, which was not formed, we found a relatively high formation of 63 (30% isolated yield, non-optimised). The exact mechanism for this deallylation is not known to us, but a suggested mechanism is presented in Scheme 57. A similar deallylation has been published previously.179

L L L L L L NMe2 L L CO Pd Pd Pd O Pd -OTf TfO Ph CPh O Ph O Ph Me2N

O L L L L O Pd Pd + Ph NMe2 Ph 63 Me2N

Scheme 57: Suggested Mechanism for the Formation of N,N-Dimethylbenzamide.

The internal arylation was also successfully applied on allyl alcohol 64, as has previously been published by Cabri,41 and with lower regioselectivity on allyl ethyl ether 65 (Scheme 58). A similar coordination between palladium and the oxygen, resulting in a five-membered intermediate, might be envisioned in these couplings, as was suggested in the case of the internal arylation of N,N-dialkylallylamines.

>99.5 88 <0.5 12 OH O 64 65 Scheme 58: Regioselectivities with Allyl Alcohol and Allyl Ethyl Ether.

49 A direct comparison between allyl alcohols and allylamines may not be appropriate. The primary allylamine does not yield any product under the present reaction conditions (see below) in contrast to the primary allyl alcohol and the regioselectivity falls when an ethyl substituent is added to the allyl alcohol, while the tertiary N,N-dimethylallylamine couples readily with excellent regioselectivity. Whether the poor regioselectivity of allyl ethyl ether is due to changes in the electron density or steric hindrance is unclear but it has been suggested that primary alcohols can be deprotonated in the reaction medium of palladium-catalysed reactions (Scheme 59).180 The deprotonated anionic allyl alcohols 66–67, as reported by Kang, may coordinate more efficiently to the palladium and this may well lead to a better regioselectivity compared to the ether, which cannot be deprotonated.

R H OH OH R H H O O Ph Pd Pd Ph I H I 66 H 67

Scheme 59: Five- or Four-Membered Intermediates Formed Through Coordination Between Palladium and an Anionic Oxygen.

Partly inspired by Blart and Ricci’s success with the internal arylation of 68122 (Scheme 60),

I H H N Me Si N Pd(OAc) , Dppb Boc + 3 Boc 2 MeO Et3N 68 MeO 69

Scheme 60: Synthesis of 69 by Internal Arylation of 68. we anticipated that a direct synthesis of internally arylated protected allylamines, such as 69 and 70 (Scheme 61), might be possible without the vinylic trimethylsilyl group of 68.

β H Pd(OAc) , dppf H 2 N γ N Ar Boc Boc K2CO3, CH3CN 80−90 oC, <20 h 69 ArOTf + 48 Pd(OAc) , dppf NPht 2 NPht K CO , DMF Ar 2 3 70 90 oC, 2−4 days

Scheme 61: Internal Arylation of Allylamides and Allylimides.

50 The ability to synthesise internally arylated allylamides and –imides through a one-step Heck-reaction would open a new and very convenient synthetic route to primary allylamine equivalents. The primary allylamine 71 (MDL-72274)181 (Scheme 62) has been evaluated as an SSAO-inhibitor,182 a new and very interesting class of compounds that recently has found applications in several fields of medicinal chemistry.183-187 A synthesis of MDL-72274 from 70 has been published.188 3-Halo-2-phenylallylamines in general are well characterised as irreversible enzyme inhibitors189-191 and 70 has been used in several other syntheses of enzyme inhibitors.188,192,193 The (E)-3-chloro substituent is vital for the selectivity for SSAO-receptors over MAO-B receptors.181 Other structural analogues of 71 have also been tested for MAO-B activity, although the compounds with the highest selectivity for the MAO-B receptor seem to be the (E)-3- fluoro-substituted compounds, such as MDL-72145 72. The latter compound was evaluated as an MAO-B inhibitor but was never marketed due to reports of haemolytic anaemia190 and was later further developed into Mofegiline 73, a drug registered for use against Parkinson’s disease (Scheme 62).

Cl F F

NH2 MeO NH2 NH2

Cl MeO F MDL-72274 MDL-72145 Mofegiline 71 72 73

Scheme 62: SSAO- and MAO-B Inhibitors.

A number of couplings between aryl triflates and Boc- (Scheme 63) and phthalimido- protected allylamines (Scheme 64) were made with thermal as well as microwave heating.

The regioselectivities were excellent with the secondary Boc-protected allylamine, especially with thermal heating, although the microwave-heated reactions also gave a very high preference for arylation at the internal carbon. One interesting result was the relatively high yield for the microwave-heated reaction with Boc-protected allylamine and 48f (entry 6, Scheme 63) as compared to the thermal reaction. The tertiary phthalimido-protected allylamines needed longer reaction times and resulted in moderate to very poor regioselectivities, although a few aryl triflates (entries 3 and 5, Scheme 64) could be coupled with good regioselectivity.

51 Thermal Heatinga Microwave Heatingb

Selectivity Isolated Time (min)/ c Isolated Entry Product c β/γ β/γ Yield 69 (%)d Effect (W) Yield 69 (%)d

NHBoc

1. MeO 69a >99.5/0.5 41 6/20 >99.5/0.5 34

NHBoc

2. OMe 69j >99.5/0.5 74 6/20 95/5 64

Me NHBoc

3. Me 69c >99.5/0.5 68 3/20 98/2 68 Me NHBoc

4. 69d >99.5/0.5 70 4/20 93/7 40 NHBoc

5. 69e >99.5/0.5 70 5/20 >99.5/0.5 56

NHBoc

6. Cl Cl 69f >99.5/0.5 7 6/35 >99.5/0.5 63

NHBoc O 7. 69g >99.5/0.5 45 5/20 >99.5/0.5 64 Me

aThe thermal reactions with Boc-protected allylamine were run in 2.5 mmol scale under nitrogen atmosphere at 80 oC (entries 3-6) or 90 °C (entries 1-2, 7) with 1 equiv of aryl triflate 48; 0.03 equiv of Pd(OAc)2; 0.132 equiv of dppf; 3 equiv of Boc-protected allylamine; 1.2 equiv of K2CO3 and 10 mL of acetonitrile. All reactions were completed within 20 h. bContinuous irradiation at 2450 MHz. The microwave heated reactions were run in 1.0 mmol scale in septum-sealed Pyrex tubes with 1 equiv of aryl triflate 48, 3 equiv of Boc-protected allylamine, 0.03 equiv of Pd(OAc)2, 0.132 equiv of dppf and 1.2 c equiv of K2CO3 in 1.0 mL of DMF. Selectivity internal β-arylation/terminal γ-arylation. Determined by GC-MS and 1H-NMR. d>95% by GC-MS.

Scheme 63: Internal Arylation of Boc-Protected Allylamine with Aryl Triflates 48.

52 Thermal Heatinga Microwave Heatingb

Isolated Selectivity Isolated Time (min)/ c Entry Product c d β/γ d β/γ Yield 70 (%) Effect (W) Yield 70 (%)

NPht

1. e e MeO 70a 57/43 18 5/20 67/33 21

NPht

2. OMe 70j 86/14 57 6/20 86/14 24

Me NPht

3. Me 70c 96/4 65 6/20 89/11 42 Me NPht

4. 70d 88/12 34 5/20 45/55 18e NPht

5. 70e 96/4 40 6/20 60/40 23

NPht

f f 6. Cl Cl 70f >99.5/0.5 9 5/20 >99.5/0.5 6

NPht

7. O 70g 50/50 17g 5/20 50/50 13g Me

aThe thermal reactions with phthalimido-protected allylamine were run in 2.5 mmol scale under nitrogen atmosphere at 90 °C with 1 equiv of aryl triflate 48; 0.03 equiv of Pd(OAc)2; 0.132 equiv of dppf; 3 equiv of phthalimido-protected allylamine; 1.2 equiv of K2CO3 and 10 mL of DMF as solvent. The reactions were completed after 2 days (entries 1-2, 7), 3 days (entries 4-6) or 4 days (entry 3). bContinuous irradiation at 2450 MHz. The microwave heated reactions were run in 1.0 mmol scale in septum-sealed Pyrex tubes with 1 equiv of aryl triflate 48, 3 equiv of phthalimido-protected allylamine, 0.03 equiv of Pd(OAc)2, 0.132 equiv of dppf and 1.2 equiv of c K2CO3 in 1.0 mL of DMF. Selectivity internal β-arylation/terminal γ-arylation. Determined by GC-MS and 1H-NMR. d>95% by GC-MS, except where otherwise indicated. e92-93% pure. Mixture of internally and terminally arylated products. f Ratio uncertain due to low yield. g91-93% pure. Mixture of internally and terminally arylated products. No elemental analysis made.

Scheme 64: Internal Arylation of Phthalimido-Protected Allylamine with Aryl Triflates 48.

53 A deprotonation of secondary amides similar to the deprotonation of alcohols (Scheme 59) could explain the very high regioselectivities attained with the secondary Boc- allylamide. An efficient coordination between the anionic nitrogen and palladium can take place as in structure 74 (Scheme 65), which would lead to the formation of the five-membered ring intermediate 75 and the internally arylated product 69 (Scheme 65). A structure similar to 75 has been suggested by Blart and Ricci in a palladium catalysed arylation, e. g. 43 (Scheme 37).

L L L L H Pd Pd LL N Ar Pd Boc H H N N Ot-Bu N Ar X Ar Ar Boc -X, -H O Ot-Bu O 74 75 69 Scheme 65: Suggested Mechanism for the Internal Arylation of Boc-Protected Allylamine.

There are no clear indications that might prove if palladium–carbonyl coordination occurs under our reaction conditions, but it might be argued that the tertiary phthalimido-protected amine, which cannot be deprotonated and thus cannot form an anionic intermediate, could coordinate to the palladium via one of the carbonyl groups as illustrated in structure 76, Scheme 66.

L L Pd L L LL NPht Ar Pd NPht Ar Pd O O -OTf N Ar OTf Ar 70 O N

O 76 77 Scheme 66: Coordination of Palladium to the Carbonyl of an Uncharged Tertiary Allylimide.

The insertion of the aryl group would then have to go through a seven-ring intermediate 77, which when compared to the five-membered 75 should be energetically unfavoured, in order to form the internally arylated product 70 (Scheme 66). However, it cannot be ruled out that the poorer regioselectivities encountered with tertiary N-protected allylamines might reflect an insertion with no palladium–heteroatom coordination at all.

54 To estimate the reactivity of secondary Boc-protected allylamines in comparison with tertiary Boc-protected allylamines, we investigated the internal arylation of the tertiary N-methyl substituted Boc-protected amine 81, yielding product 78 (Scheme 67).

Me Me OTf N N Pd(OAc)2, dppf Boc + Boc K2CO3, CH3CN 81 80 oC, <20 h 78 67% Isolated Yield Regioselectivity 89/11

Scheme 67: Internal Arylation of the Tertiary N-Methyl, Boc-Protected Allylamine.

The tertiary 81 was in pilot studies evaluated with a number of aryl triflates and the reaction times were in most cases similar to the phthalimido-protected allylamines, i. e. longer than those of the secondary 79 (Scheme 68), although a few couplings with 81 were finished over night. The regioselectivities with 81 were moderate and more in line with the phthalimido-protected allylamine 80 than with 79. This indicates that the reactivity and regioselectivity of a secondary allylamide can be profoundly changed if a third nitrogen substituent is added, possibly by preventing coordination between palladium and an anionic nitrogen as in Scheme 65.

To bring more light on the mechanism underlying the regioselectivity in the palladium- catalysed internal arylation of heteroatom-containing olefins were eleven olefins investigated in competitive couplings (Scheme 68).

In each reaction were five equivalents of two different olefins added to the reaction mixture and the relative ratio of arylation with phenyl triflate of the two olefins was measured. To ensure that the response factors for the different compounds were comparable were equimolar amounts of the arylated olefins, when the pure, isolated internally arylated olefins were obtainable, injected on GC–MS and their respective peaks measured by GC–MS integration. When pure samples were not at hand were the reaction mixtures analysed by 1H-NMR and GC–MS. The only compound that had a differing response factor was the phenylated phthalimido-protected allylamine, which had a 1.9 times higher response factor than the other phenylated olefins. The results of the competitive couplings are summarised in Scheme 69.

55 O Me Me H N N N N NH Boc Boc Me 2 O 79 80 81 82 83 Boc Me H N N SiMe3 OH Me Boc Me 84 85 86 87 88

OEt

89 Scheme 68: Olefins Evaluated in Competitive Heck-Couplings.

The secondary Boc-amide 79 has in the competitive reactions a slight advantage over the tertiary imide 80, the tertiary amide 81 and the heteroatom-lacking alkene 87 and a high reactivity compared to the tertiary amine 82. The reactivity of the phthalimido compound 80 was comparable with 81. The oxygen containing alcohol derivatives 88– 89 and the trimethylsilyl compound 86 are all slightly more reactive than 79.

79/80 60/40a 79/86 33/67 81/82 90/10

79/81 63/37 79/87 69/31 82/87 35/65

79/82 91/9 79/88 35/65 86/89 57/43

79/83-84 NR 79/89 38/62 87/89 28/72

a 79/85 >99.5/0.5 80/81 55/45 88/89 51/49

aCalculation based on a 1.9 times larger response factor for 80.

Scheme 69: Product Distribution in Competitive Arylations.

The primary 83 and secondary, methyl-substituted 84 allylamines suppress the reaction in competition with 79, giving only traces of 69 and no arylated 83 or 84 that could be detected by GC–MS analyses. Strong palladium–nitrogen coordination can thus cause catalyst poisoning with a primary or a secondary allylamine lacking electron withdrawing substituents on the nitrogen. The electron density of the amine moiety has previously been found to influence its coordinating ability.112,116 Interestingly, allyl alcohol 88 was arylated to the same extent as allyl ethyl ether 89, which may imply that other factors besides deprotonation of the oxygen influences the regioselectivity when oxygen-containing olefins are used in the Heck-reaction (Scheme 58).

56 The data suggest that there is a balance between coordination and deactivation of the palladium catalyst that must be controlled in order to ensure the formation of a Heck- coupling product. The primary allylamine 83 and the secondary N-methylallylamine 84 deactivate the catalyst, while the tertiary N,N-dimethylallylamine 82 is coupled with excellent regioselectivity, although it is at a disadvantage to the other tested olefins in competitive couplings. Even the non-coordinating 87 is favoured over 82 (65/35) in competitive phenylations. This may be explained by a partial deactivation of palladium by electron donation from the amine moiety in 82. A tight coordination of palladium to the heteroatom may result in slower ligand dissociation and a slower rate for the overall reaction.194 This deactivation may be stronger in the primary and secondary analogues 83–84 due to reduced steric hindrance.119 Furthermore, it cannot be ruled out that an anionic amine is formed when primary and secondary allylamines are used, as amine hydrogens may be more acidic when the nitrogen is coordinated to palladium,195 although no pKa values for this hydrogen dissociation have been reported.

When comparing the tertiary allylamine to the secondary or tertiary allylamides or the tertiary allylimide is the reactivity in competitive couplings higher with the amides and imide, possibly due to a weaker coordination and deactivation from the amide or imide nitrogen as the carbonyl decreases the electron density around the nitrogen as compared to the amine nitrogen. Another explanation could be a weaker coordination of a carbonyl group of the amide or imide to palladium compared to the nitrogen coordination of amines. The secondary amide 79 is more reactive in competitive couplings than the tertiary amide 81 and –imide 80 and may coordinate more efficiently to the palladium through an anionic intermediate.120 It is in this context interesting to note that the relatively electron-poor 79 is at a disadvantage to the β-stabilising 86 (33/67) in competitive experiments (Scheme 69).

A large steric bulk on the olefin may hamper the reaction as a loss of reactivity is noted when an additional Boc group is added to 79, giving the di-Boc protected 85. Couplings with 85 yield no product and the presence of 85 in a competitive coupling with 79 does not disturb the product formation of 69, indicating that the reason for the lack of reactivity of 85 is not deactivation of the catalyst (Scheme 69). The synthesis of the terminally arylated 85, which should not be as sensitive to steric hindrance, appears to run smoothly with vinyl triflates and monodentate ligands.196

In short, very high regioselectivities, which can partly be explained by Pd2+–N coordination, could be induced with N,N-dialkylallylamines and Boc-protected allylamines with thermal as well as microwave heating. Phthalimido-protected allylamines resulted in arylations with lower and less reproducible regioselectivities, especially under microwave heating.

57 5.3 Internal Arylation of Fluoroallylamines (Appendix)

As was discussed in chapter 5.2 was a synthetic route to SSAO-inhibitors, such as MDL 72274 71, open through the phthalimido-protected allylamine 70 but the (E)-3-fluoro-2- arylallylamines, such as MDL-72145 72, Scheme 62, were not accessible by the same method.

We realised that if the methodology of highly regioselective internal arylation of protected allylamines could be taken advantage of in a direct internal arylation of 90, this would result in a highly interesting and very short synthetic route to 91 and 92 (Scheme 70). F F

NR2 Pd(OAc)2, dppf NR Deprotection NH2 ArOTf + Ar 2 Ar F K CO , CH CN 90 2 3 3 91 92 R=Boc and H or Pht

Scheme 70: Envisioned Synthesis of Irreversible MAO-B Inhibitors.

We synthesised 90, largely following McCarthy’s method of producing terminally fluorinated olefins,197,198 as described in Scheme 71.

Ph O S F OEt O O O O O

EtO HCl/H2O H PhSO2CH2F N N N 78% ClP(O)(OEt)2 O 93 O 94 60% O 95

Bu Sn H 3 F F O HSnBu3 O MeONa AIBN MeOH N 41% N 65%

O O 96 90

Scheme 71: Synthesis of Terminally Fluorinated N-Allylimide Compounds.

58 Deprotection of the acetal 93 followed by a fluoro-Pummerer reaction of 94199 with 200,201 PhSO2CH2F resulted in compound 95, where the regioisomers could be separated by circular chromatography, although it was found to be unnecessary as the radical induced substitution of the PhSO2- by the Bu3Sn-group resulted in regiochemical scrambling of the vinylic fluorous atom. Separation by circular chromatography of the two regioisomers of 96 could be performed as well. The removal of the Bu3Sn-group was most easily made in refluxing NaOMe/MeOH198 although heating with CsF was also effective,198,202 yielding 90. Compounds 90 and 95 were evaluated as olefins in Heck-couplings with phenyl-, 1-naphthyl- and 2,3,5-trimethylphenyl triflates, under the same reaction conditions used for the internal arylation of protected allylamines, without any trace of the desired internally arylated products 91. The deprotection of 90 with a subsequent mono-protection with Boc was planned but abandoned when it was clear that no Heck-coupling could be discerned with 90 or 95. One explanation for the lack of success of Heck-couplings with terminally fluoro-substituted allylic compounds could be the low electron density of the olefin, which plausibly might lead to an unfavoured coordination with the cationic palladium complex and a suppressed reaction. Vinylic fluoroolefins have previously been reported to be compatible with Stille-203-205 (as in product 97)203,204 and Suzuki-coupling205,206 (98)206 conditions (Scheme 72).

F I F Pd(PPh3)2 + CuI 97 SnBu3 90% O O F (HO) B F 2 Pd(PPh3)2 + Br 98 86%

Scheme 72: Palladium-Catalysed Synthesis of Fluoroolefins.

In short, the direct internal arylation of terminally fluorinated protected allylamines failed, possibly due to an unfavoured interaction between the cationic palladium and the electron-poor olefin.

59 5.4 Microwave Heated Fluorous Stille-Couplings and Radical Reactions (Paper II)

Curran has in a series of papers taken advantage of the insolubility of fluorous compounds in standard organic and aqueous phases in separations and presented several elegant methods where the tagging of tin reagents with fluorous tails, such as 99 and 100, improved the purification process (Scheme 73).

Partitioning inorganic water

1 1 2 R Sn(CH2CH2C10F21)3 Reaction organic R R +R2X

fluorous 1 XSn(CH2CH2C10F21)3 R = Ph 99 + R1= H 100 1 R Sn(CH2CH2C10F21)3

Scheme 73: Work-Up Procedure with Fluorous Reagents.

The C10F21-tags used in 99 and 100 were introduced in fluorous Ugi- and Biginelli reactions were the normally used fluorous tags (C6F13) did not have the power to 156 partition large compounds fully to the fluorous phase. The further use of C10F21-tags in thermal reactions resulted in poor yields and irreproducible result, most likely a result of the very poor solubility of the highly fluorinated compounds.207 In this context was microwave-heated chemistry138,208 evaluated with great success in chemical reactions with compounds 99 and 100. It was shown that the high temperatures reached in the microwave vessels allowed the dissolution of 99 and 100 in either a mixture of TBF and t-BuOH (for radical-mediated reactions) or in DMF (for Stille-couplings), which was easily ascertained as the fluorous compounds were insoluble solids before microwave- heating was applied and fully dissolved after a couple of minutes in the single-mode microwave oven.

60 A number of Stille-couplings with 99 were performed in moderate to high isolated yields under microwave irradiation (Scheme 74). The experiments are summarised in Scheme 75.

Br Ph 99 Pd(PPh ) Cl , LiCl R 3 2 2 R

Scheme 74: Stille-Couplings with Highly Fluorinated Tin Reagents.

Products 101–103 were formed in moderate to high yields, while the isolated yield of 104 was lowered by formation of sideproducts, mainly diphenylmethane. The isolated yield of 102 could be raised marginally by switching to a water-soluble catalyst209,210 (entry 5, Scheme 75). Side products arising from aryl scrambling could thus be removed by extraction.

Organo Time (min)/ Isolated Entry a Product Halide Effect (W) Yield (%)b

I Ph 1. 6/50 101 78

Br Ph

2. O 6/50 O 102 71

Me Br Me Ph 3. 6/50 103 69 O2N O2N

Br Ph 4. 6/50 104 46

Br Ph 5. 6/50 c O O 102 75

Me Me

a b c Continuous irradiation at 2450 MHz. >95% by GC/MS. Pd(OAc)2 and P(m-PhSO3Na)3 instead of Pd(PPh3)2Cl2 as precatalyst.

Scheme 75: Microwave Promoted Stille-Couplings with F-21 Compound 99.

61 A number of radical-mediated reactions211 were successfully executed as well in good to excellent isolated yields (Scheme 76). Reduction of adamantyl bromide, yielding adamantane 105, and ring-closures, yielding 106 and 107 could be performed in high to excellent yields. Notably was the use of a catalytic amount of the fluorous tin hydride

100, in combination with NaBH3CN, used in the addition of adamantyl bromide to acrylonitrile, yielding the product 108. This result is interesting not only as the amount of expensive (in this case commercially unavailable) tin compound could be reduced by at least 90% but also because the amount of tin reagents, which are toxic, can be minimised. The yield of this addition could actually be increased when the insolubility of the tin-containing by-products was taken advantage of in a simplified work-up procedure, where the reaction mixture simply was filtered through a plug of silica to yield >95% pure product 108 in 77% isolated yield. The corresponding isolated yield from the standard purification procedure, with three-phase extraction and silica chromatography was 64% (Scheme 76).

Organo Time (min)/ Isolated Entry Reagents a Product Halide Effect (W) Yield (%)b

1. Br 100 5/35 H 105 81

2. 106 93 N 100 5/35 N cbz cbz

Ph Ph Ph Ph

3. Br 100 5/50 107 78

0.1 equiv 100c 4. Br 10/60 64/77d + CN CN 108

a b c Continuous irradiation at 2450 MHz. >95% by GC/MS. NaBH3CN added to recycle the tin hydride. d64% isolated yield by extraction, 77% isolated yield by filtration.

Scheme 76: Microwave Promoted Radical Reactions with F-21 Compound 100.

62 The removal of by-products of 100 by filtration at room temperature could be seen as a hybrid solid phase/liquid phase technique. The filtration allows for a work-up procedure reminiscent of solid-phase techniques at the purification step while enjoying the advantages of homogenous reactions in solution at the reaction step. The fluorous approach can thus be said to be a means of integrating synthetic and purification strategy.

The first example of a microwave heated one-pot hydrostannylation of an alkyne 109 and subsequent Stille-coupling of the resulting vinylic stannane, yielding 110, could be performed under microwave irradiation in 44% overall isolated yield (Scheme 77). This method was subsequently adopted and further developed by Maleczka.212

1. 100, AIBN, BTF O O O 10 min/60 W EtO Ph

EtO OEt 2. PhI, Pd(OAc)2, CuO, H OEt DMF/BTF 8 min/60 W O 109 E/Z 5:1 110 44%

Scheme 77: One-Pot Hydrostannylation and Stille-Coupling.

In short, the highly fluorinated tags 99 and 100, which were difficult to use under thermal heating, proved to be efficient in synthesis under microwave irradiation. The high fluorous content of the reactants could be used to improve the purification step. High yields for Stille- and radical-mediated reactions are reported, as well as a one-pot procedure for hydrostannylation and subsequent Stille-coupling of alkynes under microwave heating.

63 6 Concluding Remarks

The results of the aims of the thesis are summarised below. a) Regioselectivities that in most cases are high to excellent have been reported in novel palladium-catalysed Heck-arylation reactions with various allylic substrates. The β-stabilising effect of silicon enhances the regiocontrol in the internal arylation of allyltrimethylsilane, while a coordination between palladium and nitrogen induces very high regioselectivities in the arylation of N,N-dialkylallylamines and the Boc-protected allylamine, producing β-arylated arylethylamines, which are of interest for applications in medicinal chemistry. Phthalimido-protected allylamines are arylated with poor to moderate regioselectivity. Furthermore, considerable effort has been laid down to explain the bases for the high regioselectivities attained. The results of these studies should improve the understanding of the regiocontrol in the Heck-reaction. b) The internal arylation of protected allylamines presented a shorter and more attractive synthesis of a precursor to SSAO-inhibitors than the previously published synthetic route. The direct synthesis of MAO-B inhibitors, through the internal arylation of a terminally fluorinated, protected allylamine, failed under the present reaction conditions. c) Single-mode microwave heating has been applied on a number of Heck-, Stille- and radical-mediated reactions. The reactions were in all instances completed in three to ten minutes and the regioselectivities were in most cases comparable with those performed under thermal heating, with the exception of the microwave-heated internal arylation of phthalimido-protected allylamines, where the regiocontrol was poor. d) Fluorous chemistry with highly fluorinated tags that are inert under thermal heating could be performed with microwave heating. The insolubility of highly fluorinated compounds in ordinary organic solvents at room temperature was utilised in the rapid and efficient purification of reaction mixtures.

Corrections to the original publications:

Paper I: In the texts to Table 1 and Table 2 should the amount of dppf be 0.132 mmol.

Paper III: The numbers of the products in Table 1 should be 2.

64 7 Appendix

2-Phthalimido-1-ethanal (93). 50 g of phthalimidoacetaldehyde diethylacetal was added to 1M HCl and heated at 100 °C in 6 h. The reaction was monitored by GC–MS and the GC–MS samples were prepurified by extraction between diethyl ether and water and dried over potassium carbonate before analysis. After full conversion of the starting material had been achieved was the reaction mixture cooled to –4 °C over night and the white crystals was collected by filtration and washed with cold water. The 1 isolated yield after recrystallisation with CHCl3/isohexane was 78% and the melting point 114 °C. H 13 NMR (270 MHz, CDCl3) δ 9.65 (s, 1H), 7.90-7.74 (m, 4H), 4.56 (s, 2H); C NMR (67.8 MHz, CDCl3) δ 193.5, 167.5, 134.3, 131.9, 123.6, 47.3. MS m/z (relative intensity 70 eV) 189 (M+, 2), 160 (100), 133 (12), 104 (14), 77 (18). 1-Fluoro-1-phenylsulfonyl-3-phthalimido-1-propene (94). Fluoromethyl phenyl sulfone200 (20.4 g, 118 mmol) and diethylchlorophosphate (17.1 mL, 118 mmol) was mixed in 200 ml of dry THF under nitrogen atmosphere in a 1L three-necked round-bottom flask with magnetic stirring. The reaction mixture was cooled to –78 °C and 250 mL LiHMDS (1M in THF) was added. The reaction was stirred at –78 °C for 30 min and 93 (15.0 g, 84 mmol) dissolved in 20 mL dry THF was added and the reaction mixture was allowed to reach room temperature. The reaction was monitored with GC–MS and stirring was continued for 30 h before the aldehyde had been consumed. The reaction was quenched with saturated aq. NH4Cl and extracted with CH2Cl2, dried with MgSO4, filtrated and the solvent was removed under reduced pressure. The product 94 was purified by silica flash chromatography (eluents: isohexane/ethyl acetate 2:1) and by circular chromatography (eluents: isohexane/ethyl acetate 4:1) and 1 isolated in 60% (E/Z 4:1): H NMR (270 MHz, CDCl3) (E)-1-Fluoro-1-phenylsulfonyl-3-phthalimido-1- propene: δ 8.0-7.6 (m, 9H), 6.33 (dt, J = 7, 31 Hz, 1H), 4.49 (dd, J = 2, 7 Hz, 2H). 13C NMR (67.8 MHz, CDCl3) δ 167.5, 155.9, 151.5, 137.3, 134.9, 131.8, 129.5, 129.0, 123.5, 114.9, 32.4. (Z)-1-Fluoro-1- phenylsulfonyl-3-phthalimido-1-propene: δ 8.1-7.6 (m, 9H), 5.84 (dt, J = 7, 20 Hz, 1H), 5.05 (dd, J = 3, 7 13 Hz, 2H); C NMR (67.8 MHz, CDCl3) δ 167.5, 155.9, 151.5, 137.3, 134.9, 131.8, 129.5, 129.0, 123.5, 114.7, 32.3. MS m/z (relative intensity 70 eV) 204 (100), 175 (7), 160 (5), 147 (8), 130 (57). 1-Fluoro-3-phthalimido-1-tributylstannyl-1-propene (95). 94 (5.2 g, 15 mmol), HSnBu3 (9 mL, 33 mmol) and AIBN (150 mg) was mixed in 100 mL toluene and refluxed for 6 h. The reaction was monitored with TLC (isohexane/ethyl acetate 2:1). The reaction was not completed after 6 h, but the isolated yield did not improve with longer reaction times. The reaction mixture was extracted with CH2Cl2/water, dried over MgSO4 and the solvent was removed under reduced pressure. Excess HSnBu3 was removed by bulb-to-bulb distillation (~180 °C/1 mm Hg) and the residue was separated by circular chromatography (isohexane/ethyl acetate 10:1), where 10% unreacted starting material was recovered. 1 The (E) and (Z) isomers were isolated in 41% yield, (E/Z) 5:2. H NMR (270 MHz, CDCl3) (E)- 1- Fluoro-3-phthalimido-1-tributylstannyl-1-propene: δ 7.9-7.5 (m, 4H), 4.99 (dt, J = 7, 51 Hz, 1H), 4.43 13 (dd, J = 2, 8 Hz, 2H), 1.8-0.8 (m, 27H). C NMR (67.8 MHz, CDCl3) δ 167.8, 133.9, 132.2, 123.2, 118.7, 117.5, 31.7, 28.9, 27.2, 13.6, 9.9. (Z)- 1-Fluoro-3-phthalimido-1-tributylstannyl-1-propene: δ 7.9- 7.7 (m, 4H), 6.00 (dt, J = 8, 35 Hz, 1H), 4.16 (dd, J = 2, 8 Hz, 2H), 1.7-0.8 (m, 27H); 13C NMR (67.8 MHz, CDCl3) δ 167.7, 133.8, 132.2, 123.2, 118.6, 117.5, 31.4, 28.7, 27.1, 13.6, 9.9. 1-Fluoro-3-phthalimido-1-propene (90). 95 (81 mg, 0.16 mmol) was mixed with a fresh 1M NaOMe solution (220 µL) and dry THF (5 mL). The mixture was heated at 60 °C for 2 h and was monitored by TLC (isohexane/ethylacetate 10:1). The product was purified by circular chromatography (ethyl acetate elutes side-products, methanol elutes the product). The product 90 was isolated in 65% yield as a 5:2 1 mixture of (Z) and (E) isomers. H NMR (270 MHz, CDCl3) (Z)- 1-Fluoro-3-phthalimido-1-propene: δ 7.8-7.7 (m, 4H), 6.61 (dd, J = 85, 5 Hz, 1H), 5.07 (dddd, J = 43, 8, 8, 5 Hz, 1H), 4.07 (d, J = 8 Hz, 2H). 13 C NMR (67.8 MHz, CDCl3) δ 176.2, 152.3, 148.5, 137.7, 130.5, 108.7, 29.2. (E)- 1-Fluoro-3- phthalimido-1-tributylstannyl-1-propene: δ 7.8-7.7 (m, 4H), 6.85 (dd, J = 84, 11 Hz, 1H), 5.52 (dddd, J 13 =18, 11, 8, 8 Hz, 1H), 3.87 (d, J = 8 Hz, 2H); C NMR (67.8 MHz, CDCl3) δ 173.8, 154.5, 151.4, 137.7, 128.4, 109.4, 34.3. MS m/z (relative intensity 70 eV) 205 (100), 187 (10), 176 (17), 160 (8), 104 (22).

65 8 Supplementary Material

Paper I 2-(4-Methoxyphenyl)-3-(trimethylsilyl)-1-propene (3a).5c Compound 3a was obtained in 42% yield (3a+1a) at 60 ºC. The eluent was isohexane/diethyl ether 19:1 and the products were further purified by bulb-to-bulb distillation (∼100 ºC at 10 mm Hg). 2-(4-t-Butylphenyl)-3-(trimethylsilyl)-1-propene (3b). Compound 3b was obtained in 47% yield (3b+1b) at 60 ºC. An alumina column was used for chromatography. The eluent was isohexane and the products were further purified by bulb-to-bulb distillation (∼110 ºC at 10 mm Hg). 1 H NMR (270 MHz, CDCl3) δ 7.32 (app d, J = 1.7 Hz, 4 H), 5.13 (s, 1 H), 4.82 (m, 1 H), 2.00 (d, J = 0.7 Hz, 2 H), 1.31 (s, 9 13 H), -0.09 (s, 9 H); C NMR (67.8 MHz, CDCl3) δ 150.1, 146.2, 139.7, 125.9, 124.9, 109.3, 34.4, 31.3, + 25.9, -1.4; MS m/z (relative intensity 70 eV) 246 (M , 12), 189 (61), 73 (100). Anal. calcd for C16 H26 Si: C, 78.0; H, 10.6. Found: C, 77.9; H, 10.4. 2-(2,3,5-Trimethylphenyl)-3-(trimethylsilyl)-1-propene (3c). Compound 3c was obtained in 69% yield (3c+1c) after 10 days at 60 ºC and in 60% after 16 h at 80 ºC. The eluent was isohexane and the products 1 were further purified by bulb-to-bulb distillation (∼100 ºC at 10 mm Hg). H NMR (270 MHz, CDCl3) δ 6.87 (s, 1 H), 6.78 (s, 1 H), 4.98 (m, 1 H), 4.73 (d, J = 2.3 Hz, 1H), 2.26 (s, 3 H), 2.23 (s, 3 H), 2.19 (s, 3 13 H), 1.88 (d, J = 1.0 Hz, 2H), -0.08 (s, 9 H); C NMR (67.8 MHz, CDCl3) δ 148.5, 144.7, 136.7, 134.1, 129.7, 129.0, 126.8, 111.9, 29.1, 20.8, 20.4, 16.3, -1.7; MS m/z (relative intensity 70 eV) 232 (M+, 47), 217 (60), 73 (100). Anal. calcd for C15 H24 Si: C, 77.5; H, 10.4. Found: C, 77.4; H, 10.0. 2-Phenyl-3-(trimethylsilyl)-1-propene (3d).5c Compound 3d was obtained in 67% yield (3d+1d) at 60 ºC. The eluent was isohexane and the products were further purified by bulb-to-bulb distillation (∼105 ºC at 10 mm Hg). 2-(1-Naphthyl)-3-(trimetylsilyl)-1-propene (3e).5c Compound 3e was obtained in 77% yield (3e+1e) at 60 ºC. The eluent was isohexane. No bulb-to-bulb distillation was needed. 2-(4-Acetylphenyl)-3-(trimethylsilyl)-1-propene (3g). Compound 3g was obtained in 31% yield (3g+1g) at 80 ºC. The eluent was isohexane/diethyl ether 9:1 and the products were further purified by 1 bulb-to-bulb distillation (∼135 ºC at 6 mm Hg). H NMR (270 MHz, CDCl3 ) δ 7.89 (m, 2 H), 7.47 (m, 2 H), 5.21 (d, J = 1.3 Hz, 1 H), 4.96 (dd, J = 1.0 Hz, 1.3 Hz, 1 H), 2.81 (s, 3 H), 2.03 (d, J = 1.0 Hz, 2 H), - 13 0.13 (s, 9 H); C NMR (67.8 MHz, CDCl3) δ 197.6, 147.5, 145.7, 135.8, 128.3, 126.4, 112.1, 26.5, 25.9, + -1.5; MS m/z (relative intensity 70 eV) 232 (M , 65), 217 (8), 73 (100). Anal. calcd for C14H20OSi: C, 72.4; H, 8.7. Found: C, 72.7; H, 8.6. 2-(4-Cyanophenyl)-3-(trimethylsilyl)-1-propene (3h). Compound 3h was obtained in 59% yield (3h+1h) at 80 ºC. The eluent was isohexane/diethyl ether 19:1 and the products were further purified by 1 bulb-to-bulb distillation (∼115 ºC at 10 mm Hg). H NMR (270 MHz, CDCl3 ) δ 7.58 (m, 2 H), 7.47 (m, 2 H), 5.20 (d, J = 1.0 Hz, 1 H), 4.99 (dd, J = 1.0 Hz, 1.0 Hz, 1 H), 2.00 (d, J = 1.0 Hz, 2 H), -0.11 (s, 9 H); 13 C NMR (67.8 MHz, CDCl3) δ 147.3, 145.1, 131.9, 126.8, 118.9, 112.9, 110.7, 25.4, -1.5; MS m/z + (relative intensity 70 eV) 215 (M , 32), 200 (7), 73 (100). Anal. calcd for C13H17NSi: C, 72.5; H, 8.0; N, 6.5. Found: C, 72.7; H, 7.9; N, 6.7.

Paper III

Entry 5, Table 1 Entry 6, Table 1 250 200 250 150 200 150 100 100 Degrees Degrees 50 Centigrade Degrees 50 0 Centigrade 0 0 60 120 180 240 300 360 420 480 0 60 120 180 240 300 360 420

Seconds Seconds

66 2-(3,4-Dimethoxyphenyl)-3-N,N-dimethylamino-1-propene (2a). Compound 2a was isolated in 41% yield at 80 °C and 35% yield after microwave irradiation (3 min/20 W). The boiling point at bulb-to-bulb 1 distillation was ~115 °C at 5 mm Hg. H NMR (270 MHz, CDCl3) δ 7.09-7.05 (m, 2H), 6.85-6.81 (m, 1H), 5.37 (d, J = 2 Hz, 1H), 5.14 (q, J = 1 Hz, 1H), 3.90 (s, 3H), 3.87 (s, 3H), 3.26 (d, J = 1 Hz, 2H), 2.24 13 (s, 6H); C NMR (67.8 MHz, CDCl3) δ 148.7, 148.6, 144.7, 133.0, 118.6, 114.0, 110.9, 109.5, 64.7, 55.84, 55.76. 45.3. MS m/z (relative intensity 70 eV) 221 (M+, 2), 206 (0.4), 178 (25), 58 (100). Anal. Calcd for C13H19NO2: C, 70.6; H, 8.65; N, 6.33. Found: C, 70.5; H, 8.8; N, 6.1. 3-N,N-Dimethylamino-2-(4-methoxyphenyl)-1-propene (2b).3a Compound 2b was isolated in 40% yield at 80 °C and 30% yield after microwave irradiation (5 min/20 W). The boiling point at bulb-to-bulb distillation was ~115 °C at 8 mm Hg. 3-N,N-Dimethylamino-2-(2-methoxyphenyl)-1-propene (2c). Compound 2c was isolated in 49% yield at 80 °C and 46% yield after microwave irradiation (5 min/20 W). The boiling point at bulb-to-bulb 1 distillation was ~105 °C at 6 mm Hg. H NMR (270 MHz, CDCl3) δ 7.29-7.18 (m, 2H), 6.96-6.86 (m, 2H), 5.33 (m, 1H), 5.19 (d, J = 2 Hz, 1H), 3.83 (s, 3H), 3.32 (s, 2H), 2.21 (s, 6H); 13C NMR (67.8 MHz, CDCl3) δ 156.7, 146.0, 130.9, 130.1, 128.5, 120.6, 117.0, 110.7, 64.9, 55.3, 45.4. MS m/z (relative + intensity 70 eV) 191 (M , 3), 176 (1), 160 (3), 146 (5), 58 (100). Anal. Calcd for C12H17NO: C, 75.4; H, 8.96; N, 7.32. Found: C, 75.2; H, 9.2; N, 7.3. 2-(4-t-Butylphenyl)-3-N,N-dimethylamino-1-propene (2d). Compound 2d was isolated in 35% yield at 80 °C and 21% yield after microwave irradiation (5 min/15 W). The boiling point at bulb-to-bulb distillation was ~105 °C at 8 mm Hg. Further purification of contaminating 2f was made by applying 1 reduced pressure. H NMR (270 MHz, CDCl3) δ 7.45 (m, 2H), 7.36 (m, 2H), 5.45 (d, J = 2 Hz, 1H), 5.19 13 (q, J = 1 Hz, 1H), 3.28 (s, 2H), 2.25 (s, 6H), 1.32 (s, 9H); C NMR (67.8 MHz, CDCl3) δ 150.3, 144.8, 137.0, 125.7, 125.2, 114.6, 64.5, 45.4, 34.4, 31.3. MS m/z (relative intensity 70 eV) 217 (M+, 10), 174 (31), 159 (9), 128 (7), 58 (100). Anal. Calcd for C15H23N: C, 82.9; H, 10.7; N, 6.44. Found: C, 82.9; H, 10.8; N, 6.4. 3-N,N-Dimethylamino-2-(2,3,5-trimethylphenyl)-1-propene (2e). Compound 2e was isolated in 81% 1 yield at 80 °C and 48% yield after microwave irradiation (5 min/15 W). H NMR (270 MHz, CDCl3) δ 6.90 (s, 1H), 6.80 (s, 1H), 5.40 (m, 1H), 5.03 (m, 1H), 3.10 (m, 2H), 2.26 (s, 6H), 2.23 (s, 3H), 2.15 (s, 13 3H); C NMR (67.8 MHz, CDCl3) δ 147.4, 141.8, 136.8, 134.3, 130.2, 129.6, 126.8, 115.6, 66.1, 45.8, 20.8, 20.4, 16.0. MS m/z (relative intensity 70 eV) 203 (M+, 8), 158 (13), 129 (5),128 (5), 58 (100). Anal. Calcd for C14H21N: C, 82.7; H, 10.4; N, 6.9. Found: C, 82.8; H, 10.2; N, 6.7. 3-N,N-Dimethylamino-2-phenyl-1-propene (2f).3a Compound 2f was isolated in 42% yield at 80 °C and 24% yield after microwave irradiation (5 min/15 W). The boiling point at bulb-to-bulb distillation was ~90 °C at 10 mm Hg. 2-(2,4-Dichlorophenyl)-3-N,N-dimethylamino-1-propene (2h). Compound 2h was isolated in 38% yield at 1 80 °C. The boiling point at bulb-to-bulb distillation was ~120 °C at 8 mm Hg. H NMR (270 MHz, CDCl3) δ 7.37 (m, 1H), 7.26-7.14 (m, 2H), 7.5 (m, 4H), 5.45 (q, J = 2 Hz, 1H), 5.17 (m, 1H), 3.23 (s, 2H), 2.22 (s, 13 6H); C NMR (67.8 MHz, CDCl3). 144.5, 139.0, 133.4, 132.8, 131.2, 129.5, 126.9, 118.9, 65.0, 45.4. MS m/z (relative intensity 70 eV) 228 (0.3), 194 (9), 115 (5), 58 (100). Anal. Calcd for C11H13Cl2N: C, 57.4; H, 5.69; N, 6.08. Found: C, 57.3; H, 5.8; N, 6.2. 2-(4-Acetylphenyl)-3-N,N-dimethylamino-1-propene (2i). Compound 2i was isolated in 48% yield at 80 °C. The boiling point at bulb-to-bulb distillation was ~130 °C at 1 mm Hg. 1H NMR (270 MHz, CDCl3) δ 7.92 (m, 2H), 7.58 (m, 2H), 5.54 (d, J = 1 Hz, 1H), 5.31 (q, J = 1 Hz, 1H), 3.29 (s, 2H), 2.58, (s, 13 3H), 2.21 (s, 6H); C NMR (67.8 MHz, CDCl3) δ 197.7, 144.7, 144.5, 136.0, 128.4, 126.4, 117.3, 64.3, 45.2, 26.6. MS m/z (relative intensity 70 eV) 203 (M+, 7), 160 (1), 145 (2), 115 (4), 58 (100). Anal. Calcd for C13H17NO: C, 76.8; H, 8.43; N, 6.89. Found: C, 76.8; H, 8.3; N, 6.9. 2-(4-Cyanophenyl)-3-N,N-dimethylamino-1-propene (2j). Compound 2j was isolated in 50% yield at 80 1 °C. The boiling point at bulb-to-bulb distillation was ~105 °C at 8 mm Hg. H NMR (270 MHz, CDCl3) δ 7.60 (s, 4H), 5.53 (d, J = 1Hz, 1H), 5.34 (q, J = 1.3 Hz, 1H), 3.27 (d, J = 1 Hz, 2H), 2.20 (s, 6H); 13C NMR (67.8 MHz, CDCl3) δ 144.5, 143.9, 132.0, 126.9, 119.0, 118.1, 110.9, 64.2, 45.1. MS m/z (relative intensity + 70 eV) 186 (M , 6), 169 (1), 140 (3), 58 (100). Anal. Calcd for C12H14N2: C, 77.4; H, 7.58; N, 15.0. Found: C, 77.2; H, 7.6; N, 14.7. 2-Phenyl-3-piperidino-1-propene (2k). Compound 2k was isolated in 50% yield at 80 °C. The boiling 1 point at bulb-to-bulb distillation was ~90 °C at 1 mm Hg. H NMR (270 MHz, CDCl3) δ 7.57-7.52 (m,

67 2H), 7.37-7.23 (m, 3H), 5.46 (d, J = 2 Hz, 1H), 5.25 (q, J = 2 Hz, 1H), 3.30 (d, J = 1Hz, 2H), 2.42 (m, 13 4H), 1.56 (m, 4H), 1.42 (m, 2H); C NMR (67.8 MHz, CDCl3) δ 144.5, 140.8, 128.0, 127.3, 126.3, 114.8, 63.7, 54.6, 26.0, 24.5. MS m/z (relative intensity 70 eV) 201 (M+, 8), 158 (1), 118 (4), 115 (7), 98 (100). Anal. Calcd for C14H19N: C, 83.5; H, 9.5; N, 7.0. Found: C, 83.2; H, 9.5; N, 6.9

Paper IV

Table 1, Product 2e

250

200

150

100

Degrees 50 Centigrade

0 0 1 00 200 300 400 Seconds

Thermal Internal Phenylation of 3-N-t-Butoxycarbonylamino-N-Methyl-1-propene 6c (Eq 3). As described under General Procedure for the Thermal Internal Arylation of 3-N-t-Butoxycarbonylamino-1- propene 6a but with 6c as olefin and 5d as aryl triflate. The reaction was completed within 18 h at 80 °C. Eluents for chromatography of 4d were isohexane–ethyl acetate 9:1. Microwave Heated Arylation of 3-Phthalimido-1-propene 6b (Table 1). To a heavy-walled, oven- dried Pyrex tube were added: 1.00 mmol of aryl triflate 5a–g; 0.03 mmol (6.7 mg) of Pd(OAc)2, 0.132 mmol (73 mg) of dppf, 3.0 mmol (0.26 g) of 3-phthalimido-1-propene (0.57 g), 1.2 mmol (173 mg) of K2CO3 and 1.0 mL of DMF bubbled with N2 for 5 minutes. Otherwise as under Microwave Heated Internal Arylation of 3-N-t-Butoxycarbonylamino-1-propene 6a (Table 1). 3-N-t-Butoxycarbonylamino-2-(4-methoxyphenyl)-1-propene (2a).122 Compound 2a was obtained in 41% yield at 90 °C and 34% yield after microwave irradiation (6 min/20 W). The boiling point at bulb-to- bulb distillation was ~160 °C at 2 mm Hg. 3-N-t-Butoxycarbonylamino-2-(2-methoxyphenyl)-1-propene (2b). Compound 2b was obtained in 74% yield at 90 °C and 64% yield after microwave irradiation (6 min/20 W). The boiling point at bulb-to- 1 bulb distillation was ~ 140 °C at 2 mm Hg. H NMR (270 MHz, CDCl3) δ 7.31-7.24 (m, 1H), 7.19-7.17 (m, 1H), 6.96-6.85 (m, 2H), 5.30 (s, 1H), 5.15 (s, 1H), 4.71 (br s, 1H), 4.12 (d, J = 6 Hz, 2H), 3.82 (s, 13 3H), 1.41 (s, 9H); C NMR (67.8 MHz, CDCl3) δ 156.5, 155.8, 145.7, 130.4, 129.5, 128.9, 120.6, 114.7, 110.5, 79.0, 55.4, 45.0, 28.3. MS m/z (relative intensity 70 eV) 263 (M+, 0.4), 207 (100), 190 (15), 121 (60), 57 (40). Anal. Calcd for C15H21NO3: C, 68.4; H, 8.04; N, 5.32. Found: C, 68.2; H, 8.0; N, 5.4. 3-N-t-Butoxycarbonylamino-2-(2,3,5-trimethylphenyl)-1-propene (2c). Compound 2c was obtained in 68% yield at 80 °C and 68% yield after microwave irradiation (3 min/20 W). The boiling point at bulb-to- 1 bulb distillation was ~ 160 °C at 2 mm Hg. H NMR (270 MHz, CDCl3) δ 6.93 (s, 1H), 6.79 (br s, 1H), 5.32 (d, J = 1.7 Hz, 1H), 4.97 (d, J = 1.7 Hz, 1H), 4.74 (br s, 1H), 3.91 (d, J = 5.3 Hz, 2H), 2.28 (s, 3H), 13 2.26 (s, 3H), 2.18 (s, 3H), 1.45 (s, 9H); C NMR (67.8 MHz, CDCl3) δ 155.8, 147.2, 140.4, 136.9, 134.5, 130.7, 129.8, 126.9, 113.1, 79.5, 46.2, 28.3, 20.7, 20.4, 15.9. MS m/z (relative intensity 70 eV) 275 (M+, 2), 219 (31), 174 (100), 158 (53), 143 (42). Anal. Calcd for C17H25NO2: C, 74.1; H, 9.15; N, 5.09. Found: C, 74.1; H, 9.0; N, 5.2. 3-N-t-Butoxycarbonylamino-2-(1-naphthyl)-1-propene (2e). Compound 2e was obtained in 70% yield at 80 °C and 56% yield after microwave irradiation (5 min/20 W). The boiling point at bulb-to-bulb 1 distillation was ~165 °C at 2 mm Hg. H NMR (270 MHz, CDCl3) δ 8.10-8.00 (m, 1H), 7.88-7.78 (m, 2H), 7.52-7.42 (m, 3H), 7.33-7.30 (m, 1H), 5.57 (s, 1H), 5.21 (s, 1H), 4.76 (br s, 1H), 4.11 (d, J = 5.6 Hz, 2H), 13 1.43 (s, 9H); C NMR (67.8 MHz, CDCl3) δ 155.8, 145.3, 138.5, 133.6, 131.4, 128.2, 127.7, 126.1, 125.8, 125.5, 125.4, 125.2, 115.3, 79.4, 46.7, 28.3. MS m/z (relative intensity 70 eV) 283 (M+, 5) , 227 (6), 183

68 (36), 166 (100), 153 (99). Anal. Calcd for C18H21NO2: C, 76.3; H, 7.47; N, 4.94. Found: C, 76.1; H, 7.5; N, 4.9. 3-N-t-Butoxycarbonylamino-2-(2,4-dichlorophenyl)-1-propene (2f). Compound 2f was obtained in 7% yield at 90 °C and 63% yield after microwave irradiation (6 min/35 W). The boiling point at bulb-to-bulb 1 distillation was ~ 175 °C at 2 mm Hg. H NMR (270 MHz, CDCl3) δ 7.39-7.38 (m, 1H), 7.23-7.13 (m, 2H), 5.41 (s, 1H), 5.11 (s, 1H), 4.70 (bs, 1H), 4.02 (d, J = 6 Hz, 2H), 1.40 (s, 9H); 13C NMR (67.8 MHz, CDCl3) δ 155.7, 144.2, 137.9, 133.8, 133.1, 131.5, 129.3, 127.0, 116.4, 79.3, 45.0, 28.3. MS m/z (relative intensity 70 eV) 247 (63), 245 (97), 210 (27), 166 (65), 57 (100). Anal. Calcd for C14H17Cl2NO2: C, 55.6; H, 5.67; N, 4.63. Found: C, 55.6; H, 5.8; N, 4.7. 2-(4-Acetylphenyl)-3-N-t-butoxycarbonylamino-1-propene (2g). Compound 2g was obtained in 45% yield at 90 °C and 64% yield after microwave irradiation (5 min/20 W). The boiling point at bulb-to-bulb 1 distillation was ~175 °C at 1 mm Hg. H NMR (270 MHz, CDCl3) δ 7.95-7.86 (m, 2H), 7.51 (d, J = 8.4 Hz, 2H), 5.53 (s, 1H), 5.34 (s, 1H), 4.66 (br s, 1H), 4.21 (d, J = 5.4 Hz, 2H), 2.60 (s, 3H), 1.43 (s, 9H); 13 C NMR (67.8 MHz, CDCl3) δ 197.8, 155.6, 144.2, 143.5, 136.5, 128.5, 126.3, 115.2, 79.4, 44.1, 28.3, 26.6. MS m/z (relative intensity 70 eV) 219 (77), 204 (44), 186 (62), 175 (55), 146 (100). Anal. Calcd for C16H21NO3: C, 69.8; H, 7.69; N, 5.09. Found: C, 70.0; H, 7.3; N, 4.9. 2-(4-Methoxyphenyl)-3-phthalimido-1-propene (3a). Compound 3a was obtained in 18% yield (mixture of internally and terminally arylated products) at 90 °C and 21% yield (mixture of internally and terminally arylated products) after microwave irradiation (5 min/20 W). Melting point: 106 °C. 1H NMR (270 MHz, CDCl3) δ 7.88-7.68 (m, 4H), 7.46-7.42 (m, 2H), 6.89-6.80 (m, 2H), 5.37 (s, 1H), 5.09 (t, J = 1 Hz, 1H), 13 4.68 (t, J = 1 Hz, 2H), 3.80 (s, 3H); C NMR (67.8 MHz, CDCl3) δ 167.9, 141.6, 133.9, 132.1, 127.7, 127.5, 123.2, 120.4, 113.7, 112.5, 55.2, 41.5. MS m/z (relative intensity 70 eV) 293 (M+, 77), 278 (4), 260 + (4), 160 (7), 133 (100). High resolution MS (EI) calcd for C18H15NO3 (M ) 293.1048, found 293.1048. Anal. Calcd for C18H15NO3: C, 73.7; H, 5.15; N, 4.78. Found: C, 73.2; H, 5.0; N, 4.2. 3-Phthalimido-2-(2,3,5-trimethylphenyl)-1-propene (3c). Compound 3c was obtained in 65% yield at 90 °C and 42% yield after microwave irradiation (6 min/20 W). Melting point: 152 °C. 1H NMR (270 MHz, CDCl3) δ 7.90-7.70 (m, 4H), 6.97-8.85 (m, 2H), 5.08 (s, 1H), 4.98 (s, 1H), 4.43 (s, 2H), 2.26 (m, 13 9H); C NMR (67.8 MHz, CDCl3) δ 167.9, 144.2, 139.6, 137.0, 134.6, 134.0, 132.0, 131.0, 130.1, 127.1, 123.4, 113.1, 43.0, 20.7, 20.4, 15.8. MS m/z (relative intensity 70 eV) 305 (M+, 78), 290 (100), 262 (22), 158 (87), 143 (59). Anal. Calcd for C20H19NO2: C, 78.7; H, 6.27; N, 4.59. Found: C, 78.5; H, 6.4; N, 4.5. 2-Phenyl-3-phthalimido-1-propene (3d).188 Compound 3d was obtained in 34% yield at 90 °C and 18% yield (mixture of internally and terminally arylated products) after microwave irradiation (5 min/20 W). 2-(1-Naphthyl)-3-phthalimido-1-propene (3e). Compound 3e was obtained in 40% yield at 90 °C and 1 23% yield after microwave irradiation (6 min/20 W). Melting point: 115 °C. H NMR (270 MHz, CDCl3) δ 8.27-8.23 (m, 1H), 7.90-7.69 (m, 6H), 7.58-7.39 (m, 4H), 5.36 (t, J = 1 Hz, 1H), 5.24 (m, 1H), 4.65 (t, J = 2 13 Hz, 2H) ; C NMR (67.8 MHz, CDCl3) δ 167.9, 142.4, 137.7, 134.0, 133.6, 132.0, 131.4, 128.2, 128.0, 126.3, 125.9, 125.8, 125.5, 125.2, 123.4, 115.3, 43.4. MS m/z (relative intensity 70 eV) 313 (M+, 12), 166 (100), 152 (38), 104 (5), 77 (9). Anal. Calcd for C21H15NO2: C, 80.5; H, 4.82; N, 4.47. Found: C, 80.3; H, 5.1; N, 4.4. 2-(4-Acetylphenyl)-3-phthalimido-1-propene (3g). Compound 3g was obtained in 17% yield (mixture of internally and terminally arylated products) at 90 °C and 13% yield (mixture of internally and terminally arylated products) after microwave irradiation (5 min/20 W). Melting point: 135 °C. 1H NMR (270 MHz, CDCl3) δ 7.93-7.68 (m, 6H), 7.60-7.39 (m, 2H), 5.54 (s, 1H), 5.13 (t, J = 1 Hz, 1H), 4.72 (s, 13 2H), 2.57 (s, 3H) ; C NMR (67.8 MHz, CDCl3) δ 197.5, 167.8, 140.8, 136.5, 134.0, 132.0, 128.6, 126.6, 125.7, 123.3, 116.3, 39.5, 26.6. MS m/z (relative intensity 70 eV) 305 (M+, 94), 290 (100), 272 (5), 160 (37), 115 (16). Anal. Calcd for C19H15NO3: C, 74.7; H, 4.95; N, 4.59. Found: C, 74.4; H, 5.0; N, 4.4.

69 9 References

(1) Heck, R. F. Org. React. 1982, 27, 345-390. (2) Heck, R. F. In Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford, 1991; Vol. 4, p 833-863. (3) Cabri, W.; Candiani, I. Acc. Chem. Res. 1995, 28, 2-7. (4) Tsuji, J. In Palladium Reagents and Catalysts; John Wiley & Sons: Chichester, 1995, p 125-168. (5) Jeffery, T. In Advances in Metal-Organic Chemistry; Liebeskind, L. S., Ed.; Jai Press, Inc: Greenwich, 1996; Vol. 5, p 153-259. (6) Amatore, C.; Jutand, A. Acc. Chem. Res. 2000, 33, 314-321. (7) Beletskaya, I. P.; Cheprakov, A. V. Chem. Rev. 2000, 100, 3009-3066. (8) Larhed, M.; Hallberg, A. In Handbook of Organo-Palladium Chemistry for Organic Synthesis; Negishi, E.- i., Ed.; Wiley-Interscience: In press. (9) Negishi, E.-i. J. Organomet. Chem. 1999, 576, xv-xvi. (10) Echavarren, A. M.; Stille, J. K. J. Am. Chem. Soc. 1987, 109, 5478-5486. (11) Mitchell, T. N. Synthesis 1992, 803-815. (12) Martin, A. R.; Yang, Y. Acta Chem. Scand. 1993, 47, 221-230. (13) Suzuki, A. Pure Appl. Chem. 1994, 66, 213-222. (14) Moritani, I.; Fujiwara, Y. Tetrahedron Lett. 1967, 1119-1122. (15) Fujiwara, Y.; Moritani, I.; Matsuda, M. Tetrahedron 1968, 24, 4819-4824. (16) Heck, R. F. J. Am. Chem. Soc. 1968, 90, 5518-5526. (17) Heck, R. F. J. Am. Chem. Soc. 1969, 91, 6707-6714. (18) Farina, V. Pure Appl. Chem. 1996, 68, 73-78. (19) Ozawa, F.; Kubo, A.; Hayashi, T. Chem. Lett. 1992, 2177-2180. (20) Amatore, C.; Jutand, A.; M´Barki, M. A. Organometallics 1992, 11, 3009-3013. (21) Csákai, Z.; Skoda-Földes, R.; Kollár, L. Inorg. Chim. Acta 1999, 286, 93-97. (22) McCrindle, R.; Ferguson, G.; Arsenault, G. J.; McAlees, A. J. J. Chem. Soc., Chem. Commun 1983, 571- 573. (23) Heck, R. F.; Nolley Jr, J. P. J. Org. Chem. 1972, 37, 2320-2322. (24) Trost, B. M.; Murphy, D. J. Organometallics 1985, 4, 1143-1145. (25) Carlström, A.-S.; Frejd, T. Acta Chem. Scand. 1992, 46, 163-171. (26) Amatore, C.; Carré, E.; Jutand, A.; M'Barki, M. A. Organometallics 1995, 14, 1818-1826. (27) Ritter, K. Synthesis 1993, 735-762. (28) Rottländer, M.; Knochel, P. J. Org. Chem. 1998, 63, 203-208. (29) Fitton, P.; Rick, E. A. J. Organomet. Chem. 1971, 28, 287-291. (30) Kamikawa, T.; Hayashi, T. Tetrahedron Lett. 1997, 38, 7087-7090. (31) Reetz, M. T.; Lohmer, G.; Schwickardi, R. Angew. Chem., Int. Ed. Engl. 1998, 37, 481-483. (32) Hamann, B. C.; Hartwig, J. F. J. Am. Chem. Soc. 1998, 120, 7369-7370. (33) Littke, A. F.; Fu, G. C. J. Org. Chem. 1999, 64, 10-11. (34) Huang, J.; Grasa, G.; Nolan, S. P. Org. Lett. 1999, 1, 1307-1309. (35) Huang, J.; Nolan, S. P. J. Am. Chem. Soc. 1999, 121, 9889-9890. (36) Zhang, C.; Huang, J.; Trudell, M. L.; Nolan, S. P. J. Org. Chem. 1999, 64, 3804-3805. (37) Bräse, S.; de Meijere, A. In Metal-Catalyzed Cross Coupling Reactions; Diederich, F., Stang, P. J., Eds.; Wiley: Weinheim, 1998, p 99-166. (38) Alcazar-Roman, L. M.; Hartwig, J. F.; Rheingold, A. L.; Liable-Sands, L. M.; Guzei, I. A. J. Am. Chem. Soc. 2000, 122, 4618-4630. (39) Thorn, D. L.; Hoffmann, R. J. Am. Chem. Soc. 1978, 100, 2079-2090. (40) Samsel, E. G.; Norton, J. R. J. Am. Chem. Soc. 1984, 106, 5505-5512. (41) Cabri, W.; Candiani, I.; Bedeschi, A.; Santi, R. J. Org. Chem. 1992, 57, 3558-3563. (42) Andersson, C.-M.; Hallberg, A.; Daves Jr., G. D. J. Org. Chem. 1987, 52, 3529-3536. (43) Cabri, W.; Candiani, I.; Bedeschi, A.; Penco, S.; Santi, R. J. Org. Chem. 1992, 57, 1481-1486. (44) Andersson, C.-M.; Hallberg, A. J. Org. Chem. 1988, 53, 2112-2114. (45) Jutand, A.; Mosleh, A. Organometallics 1995, 14, 1810-1817. (46) Karabelas, K.; Westerlund, C.; Hallberg, A. J. Org. Chem. 1985, 50, 3896-3900. (47) Karabelas, K.; Hallberg, A. J. Org. Chem. 1986, 51, 5286-5290. (48) Abelman, M. M.; Oh, T.; Overman, L. E. J. Org. Chem. 1987, 52, 4130-4133. (49) Yamamoto, A. J. Organomet. Chem. 1995, 500, 337-348. (50) Amatore, C.; Carré, E.; Jutand, A. Acta Chem. Scand. 1998, 52, 100-106. (51) Ashimori, A.; Bachand, B.; Calter, M. A.; Govek, S. P.; Overman, L. E.; Poon, D. J. J. Am. Chem. Soc. 1998, 120, 6488-6499. (52) Deeth, R. J.; Smith, A.; Hii, K. K.; Brown, J. M. Tetrahedron 1998, 39, 3229-3232.

70 (53) McGuinness, D. S.; Cavell, K. J.; Skelton, B. W.; White, A. H. Organometallics 1999, 18, 1596-1605. (54) van Leeuwen, P. W. N. M.; Kamer, P. C. J.; Reek, J. N. H.; Dierkes, P. Chem. Rev. 2000, 100, 2741-2770. (55) Gan, K. S.; Hor, T. S. In Ferrocenes; Togni, A., Hayashi, T., Eds.; VCH: Weinheim, 1995, p 3-104. (56) Hayashi, T.; Konishi, M.; Kobori, Y.; Kumada, M.; Higushi, T.; Hirotsu, K. J. Am. Chem. Soc. 1984, 106, 158-163. (57) Butler, I. R.; Cullen, W. R.; Kim, T.-J.; Rettig, S. J.; Trotter, J. Organometallics 1985, 4, 972-980. (58) Tolman, C. A. Chem. Rev. 1977, 77, 313-348. (59) Hartwig, J. F.; Richards, S.; Baranano, D.; Paul, F. J. Am. Chem. Soc. 1996, 118, 3626-3633. (60) Boyes, A. L.; Butler, I. R.; Quayle, S. C. Tetrahedron Lett. 1998, 39, 7763-7766. (61) Ozawa, F.; Kubo, A.; Matsumoto, Y.; Hayashi, T.; Nishioka, E.; Yanagi, K.; Morigushi, K. Organometallics 1993, 12, 4188-4196. (62) Amatore, C.; Broeker, G.; Jutand, A.; Khalil, F. J. Am. Chem. Soc. 1997, 119, 5176-5185. (63) Shaw, B. L. New J. Chem. 1998, 77-79. (64) Herrmann, W. A.; Brossmer, C.; Öfele, K.; Reisinger, C.-P.; Priermeier, T.; Beller, M.; Fischer, H. Angew. Chem., Int. Ed. Engl. 1995, 34, 1844-1849. (65) Estevan, F.; Lahuerta, P.; Latorre, J.; Sanchez, A.; Sieiro, C. Polyhedron 1987, 6, 473-478. (66) Sato, T.; Chono, N.; Ishabashi, H.; Ikeda, M. J. Chem. Soc., Perkin Trans. 1 1995, 1115-1120. (67) Goodson, F. E.; Wallow, T. I.; Novak, B. M. J. Am. Chem. Soc. 1997, 119, 12441-12453. (68) Herrmann, W. A.; Brossmer, C.; Öfele, K.; Beller, M.; Fischer, H. J. Organomet. Chem. 1995, 491, C1-C4. (69) Morita, D. K.; Stille, J. K.; Norton, J. R. J. Am. Chem. Soc. 1995, 117, 8576-8581. (70) Kong, K.-C.; Cheng, C.-H. J. Am. Chem. Soc. 1991, 113, 6313-6315. (71) Sakamoto, M.; Shimizu, I.; Yamamoto, A. Chem. Lett. 1995, 1101-1102. (72) Segelstein, B. E.; Butler, T. W.; Chenard, B. L. J. Org. Chem. 1995, 60, 12-13. (73) Grushin, V. V. Organometallics 2000, 19, 1888-1900. (74) Brown, J. M.; Hii, K. K. Angew. Chem., Int. Ed. Engl 1996, 35, 657-659. (75) Cabri, W.; Candiani, I.; DeBernardinis, S.; Francalanci, F.; Penco, S.; Santi, R. J. Org. Chem 1991, 56, 5796-5800. (76) Kelkar, A. A.; Hanaoka, T.-a.; Kubota, Y.; Sugi, Y. J. Molecul. Cat. 1994, 88, L113-L116. (77) Saá, J. M.; Dopico, M.; Martorell, G.; García- Raso, A. J. Org. Chem. 1990, 55, 991-995. (78) Crisp, G. T.; Gebauer, M. G. Tetrahedron 1996, 52, 12465-12474. (79) Chan, T. H.; Fleming, I. Synthesis 1979, 761-786. (80) Daves Jr., G. D.; Hallberg, A. Chem. Rev. 1989, 89, 1433-1445. (81) Horn, K. A. Chem. Rev. 1995, 95, 1317-1350. (82) Reed, C. A. Acc. Chem. Res. 1998, 31, 325-332. (83) Lambert, J. B. Tetrahedron 1990, 46, 2677-2689. (84) Weber, W. P. Silicon Reagents for Organic Synthesis; Springer Verlag: New York, 1983. (85) Yamamoto, Y.; Asao, N. Chem. Rev. 1993, 93, 2207-2293. (86) Wierschke, S. G.; Chandrasekhar, J.; Jorgensen, W. L. J. Am. Chem. Soc. 1985, 107, 1496-1500. (87) Hwu, J. R.; Chen, B.-L.; Shiao, S.-S. J. Org. Chem. 1995, 60, 2448-2455. (88) Lambert, J. B.; Zhao, Y.; Emblidge, R. W.; Salvador, L. A.; Liu, X.; So, J.-H.; Chelius, E. C. Acc. Chem. Res. 1999, 32, 183-190. (89) Lambert, J. B.; Wang, G.-t.; Finzel, R. B.; Teramura, D. H. J. Am. Chem. Soc. 1987, 109, 7838-7845. (90) Lambert, J. B.; Liu, X. J. Organomet. Chem. 1996, 521, 203-210. (91) Lambert, J. B.; Emblidge, R. W.; Zhao, Y. J. Org. Chem. 1994, 59, 5397-5403. (92) Hagen, G.; Mayr, H. J. Am. Chem. Soc. 1991, 113, 4954-4961. (93) Mishima, M.; Kang, C. H.; Fujio, M.; Tsuno, Y. Chem. Lett. 1992, 2439-2442. (94) Chiavarino, B.; Crestoni, M. E.; Fornarini, S. J. Am. Chem. Soc. 1998, 120, 1523-1527. (95) Lambert, J. B.; Zhao, Y.; Wu, H. J. Org. Chem. 1999, 64, 2729-2736. (96) Davis, D. D.; Jacocks, H. M. J. Organomet. Chem. 1981, 206, 33-47. (97) Lambert, J. B.; Wang, G.-t.; Teramura, D. H. J. Org. Chem. 1988, 53, 5422-5428. (98) Apeloig, Y.; Arad, D. J. Am. Chem. Soc. 1985, 107, 5285-5286. (99) Yoshida, J.-i.; Murata, T.; Isoe, S. J. Organomet. Chem 1988, 345, C23-C27. (100) Gabelica, V.; Kresge, A. J. J. Am. Chem. Soc. 1996, 118, 3838-3841. (101) Lambert, J. B.; Ciro, S. M. J. Org. Chem. 1996, 61, 1940-1945. (102) Cope, A. C.; Kliegman, J. M.; Friedrich, E. C. J. Am. Chem. Soc. 1967, 89, 287-291. (103) Holton, R. A.; Kjonaas, R. A. J. Am. Chem. Soc. 1977, 99, 4177-4179. (104) Holton, R. A. J. Am. Chem. Soc. 1977, 99, 8083-8085. (105) Parker, K. A.; O'Fee, R. P. J. Org. Chem. 1983, 48, 1547-1550. (106) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. In Principles and Applications of Organotransition Metallic Chemistry; University Science Books: 1987, p 825-859. (107) Andersson, C.-M.; Larsson, J.; Hallberg, A. J. Org. Chem. 1990, 55, 5757-5761. (108) Larhed, M.; Andersson, C.-M.; Hallberg, A. Acta Chem. Scand. 1993, 47, 212-217.

71 (109) McCrindle, R.; Alyea, E. C.; Dias, S. A.; McAlees, A. J. J. Chem. Soc., Dalton 1979, 640-647. (110) McCrindle, R.; Ferguson, G.; Khan, M. A.; McAlees, A. J.; Ruhl, B. L. J. Chem. Soc., Dalton 1981, 986- 991. (111) Nilsson, K.; Hallberg, A. J. Org. Chem. 1990, 55, 2464-2468. (112) Albéniz, A. C.; Espinet, P.; Lin, Y.-S. Organometallics 1996, 15, 5010-5017. (113) Babar, V. J.; Khasnis, D. V.; Shinde, V. M. J. Indian Chem. Soc. 1981, 58, 970-972. (114) Lopez, G.; Garcia, G.; Sanchez, G.; Vallejo, M. C. J. Organomet. Chem. 1984, 263, 247-252. (115) Filippini, L.; Gusmeroli, M.; Riva, R. Tetrahedron Lett. 1993, 34, 1643-1646. (116) Nilsson, K.; Hallberg, A. J. Org. Chem. 1992, 57, 4015-4017. (117) Crisp, G. T.; Gebauer, M. G. Tetrahedron Lett. 1995, 36, 3389-3392. (118) Mecking, S.; Johnson, L. K.; Wang, L.; Brookhart, M. J. Am. Chem. Soc. 1998, 120, 888-899. (119) Seligson, A. L.; Trogler, W. C. J. Am. Chem. Soc. 1991, 113, 2520-2527. (120) Sigel, H.; Martin, R. B. Chem. Rev. 1982, 82, 385-426. (121) Wieber, G. M.; Hegedus, L. S.; Åkermark, B.; Michalson, E. T. J. Org. Chem. 1989, 65, 4649-4653. (122) Alvisi, D.; Blart, E.; Bonini, B. F.; Mazzanti, G.; Ricci, A.; Zani, P. J. Org. Chem. 1996, 61, 7139-7146. (123) Hay, L. A.; Koenig, T. M.; Ginah, F. O.; Copp, J. D.; Mitchell, D. J. Org. Chem. 1998, 63, 5050-5058. (124) Catellani, N.; Chiusoli, G. P.; Costa, M. J. Organomet. Chem. 1995, 500, 69-80. (125) Stone-Elander, S.; Elander, N. Appl. Radiat. Isot. 1991, 42, 885-887. (126) Elander, N.; Jones, J. R.; Lu, S.-Y.; Stone-Elander, S. Chem. Soc. Rev. 2000, 29, 239-249. (127) Loupy, A.; Petit, A.; Hamelin, J.; Texier-Boullet, F.; Jacquault, P.; Mathé, D. Synthesis 1998, 1213-1234. (128) Cotterill, I. C.; Usyatinsky, Y. A.; Arnold, J. M.; Clark, D. S.; Dordick, J. S.; Michels, P. C.; Khmelnitsky, Y. L. Tetrahedron Lett. 1998, 39, 1117-1120. (129) Strauss, C. R. Aust. J. Chem. 1999, 52, 83-96. (130) Mingos, D. M. P.; Baghurst, D. R. In Microwave Enhanced Chemistry; Kingston, H. M., Haswell, S. J., Eds.; American Chemical Society: Washington DC, 1997, p 3-53. (131) Gabriel, C.; Gabriel, S.; Grant, E. H.; Halstead, B. S. J.; Mingos, D. M. P. Chem. Soc. Rev. 1998, 27, 213- 223. (132) Lekovsek, S.; Smidovnik, A.; Koloini, T. J. Org. Chem. 1994, 59, 7433-7436. (133) Deshayes, S.; Liagre, M.; Loupy, A.; Luche, J.-L.; Petit, A. Tetrahedron 1999, 55, 10851-10870. (134) Galema, S. A. Chem. Soc. Rev 1997, 26, 233-238. (135) Stuerga, D. A. C.; Gaillard, P. J. Microwave Power Electromag. Energy 1996, 31, 101-113. (136) Langa, F.; De La Cruz, P.; De La Hoz, A.; Díaz-Ortiz, A.; Díez-Barra, E. Contemp. Org. Synth. 1997, 373- 386. (137) Hájek, M. Collect. Czech. Chem. Commun. 1997, 62, 347-354. (138) Larhed, M.; Hallberg, A. J. Org. Chem. 1996, 61, 9582-9584. (139) Hillers, S.; Sartori, S.; Reiser, O. J. Am. Chem. Soc. 1996, 118, 2087-2088. (140) Alterman, M.; Hallberg, A. J. Org. Chem. 2000, 65, 7984-7989. (141) Kaiser, N.-F. K.; Bremberg, U.; Larhed, M.; Moberg, C.; Hallberg, A. J. Organomet. Chem. 2000, 603, 2-5. (142) Gedye, R. N.; Wei, J. B. Can. J. Chem. 1998, 76, 525-532. (143) Zhu, D.-W. Synthesis 1993, 953-954. (144) Cavazzini, M.; Montanari, F.; Pozzi, G.; Quici, S. J. Fluorine Chem. 1999, 94, 183-193. (145) Horváth, I. T.; Rábai, J. Science 1994, 266, 72-75. (146) Horváth, I. T. Acc. Chem. Res. 1998, 31, 641-650. (147) Studer, A.; Hadida, S.; Ferritto, R.; Kim, S.-Y.; Jeger, P.; Wipf, P.; Curran, D. P. Science 1997, 275, 823- 826. (148) Studer, A.; Curran, D. P. Tetrahedron 1997, 53, 6681-6696. (149) Curran, D. P. Angew. Chem., Int. Ed. Engl. 1998, 37, 1174-1196. (150) Curran, D. P.; Hoshino, M. J. Org. Chem. 1996, 61, 6480-6481. (151) Curran, D. P.; Hadida, S. J. Am. Chem. Soc. 1996, 118, 2531-2532. (152) Riess, J. G.; Krafft, M. P. MRS Bulletin 1999, May 42-49. (153) Curran, D. P.; Luo, Z.; Degenkolb, P. Bioorg. Med. Chem. Lett. 1998, 8, 2403-2408. (154) Alvey, L. J.; Rutherford, D.; Juliette, J. J. J.; Gladysz, J. A. J. Org. Chem. 1998, 63, 6302-6308. (155) Juliette, J. J. J.; Rutherford, D.; Horváth, I. T.; Gladysz, J. A. J. Am. Chem. Soc. 1999, 121, 2696-2704. (156) Studer, A.; Jeger, P.; Wipf, P.; Curran, D. P. J. Org. Chem. 1997, 62, 2917-2924. (157) Linclau, B.; Sing, A. K.; Curran, D. P. J. Org. Chem. 1999, 64, 2835-2842. (158) Rocaboy, C.; Rutherford, D.; Bennet, B. L.; Gladysz, J. A. J. Phys. Org. Chem. 2000, 13, 596-603. (159) Kiss, L. E.; Rábai, J.; Varga, L.; Kövesdi, I. Synlett 1998, 1243-1245. (160) Herrera, V.; de Rege, P. J. F.; Horváth, I. T.; Husebo, T. L.; Hughes, R. P. Inorg. Chem. Commun. 1 1998, 197-199. (161) Horváth, I. T.; Kiss, G.; Cook, R. A.; Bond, J. E.; Stevens, P. A.; Rábai, J.; Mozeleski, E. J. J. Am. Chem. Soc. 1998, 120, 3133-3143. (162) Curran, D. P.; Hadida, S.; Kim, S.-Y.; Luo, Z. J. Am. Chem. Soc. 1999, 121, 6607-6615.

72 (163) Ryu, I.; Niguma, T.; Minakata, S.; Komatsu, M.; Hadida, S.; Curran, D. P. Tetrahedron Lett. 1997, 38, 7883-7886. (164) Horner, J. H.; Martinez, F. N.; Newcomb, M.; Hadida, S.; Curran, D. P. Tetrahedron Lett. 1997, 38, 2783- 2786. (165) Ryu, I.; Niguma, T.; Minakata, S.; Komatsu, M.; Luo, Z.; Curran, D. P. Tetrahedron Lett. 1999, 40, 2367- 2370. (166) Ogawa, A.; Curran, D. P. J. Org. Chem. 1997, 62, 450-451. (167) Curran, D. P.; Hadida, S.; He, M. J. Org. Chem. 1997, 62, 6714-6715. (168) Kainz, S.; Luo, Z.; Curran, D. P.; Leitner, W. Synthesis 1998, 1425-1427. (169) Curran, D. P.; Luo, Z. J. Am. Chem. Soc. 1999, 121, 9069-9072. (170) Crich, D.; Sun, S. J. Org. Chem. 1996, 61, 7200-7201. (171) Stork, G.; Sher, P. M. J. Am. Chem. Soc. 1986, 108, 303-304. (172) Cabri, W.; De Bernardinis, S.; Francalanci, F.; Penco, S.; Santi, R. J. Org. Chem. 1990, 55, 350-353. (173) Jutand, A.; Mosleh, A. J. Org. Chem. 1997, 62, 261-274. (174) Jeffery, T. Tetrahedron Lett. 2000, 41, 8445-8449. (175) Desponds, O.; Schlosser, M. J. Organomet. Chem. 1991, 409, 93-101. (176) Malek, N. J.; Moorman, A. E. J. Org. Chem. 1982, 47, 5395-5397. (177) Dimmock, J. R.; Patil, S. A.; Leek, D. M.; Warrington, R. C.; Fang, W. D. Eur. J. Med. Chem. 1987, 22, 545-551. (178) Traxter, P.; Trinks, U.; Buchdunger, E.; Mett, H.; Meyer, T.; Müller, M.; Regensass, U.; Rösel, J.; Lydon, N. J. Med. Chem. 1995, 38, 2441-2448. (179) Mori, M.; Ban, Y. Chem. Pharm. Bull. 1976, 24, 1992-1999. (180) Kang, S.-K.; Jung, K.-Y.; Park, C.-H.; Namkoong, E.-Y.; Kim, T.-H. Tetrahedron Lett. 1995, 36, 6287- 6290. (181) Lyles, G. A.; Marshall, C. M.; McDonald, I. A.; Bey, P.; Palfreyman, M. G. Biochem. Pharmacol. 1987, 36, 2847-2853. (182) Ekblom, J. Pharmacol. Res. 1998, 37, 87-92. (183) Yu, P. H.; Deng, Y. L. Atherosclerosis 1998, 140, 357-363. (184) Garpenstrand, H.; Ekblom, J.; von Arbin, M.; Oreland, L.; Murray, V. Eur. Neurol. 1999, 41, 20-23. (185) Garpenstrand, H.; Ekblom, J.; Backlund, J. B.; Oreland, L.; Rosenqvist, U. Diabet. Med. 1999, 16, 514-521. (186) Meszaros, Z.; Szombathy, T.; Karadi, I.; Romics, L.; Magyar, K. Metabolism 1999, 48, 113-117. (187) Conklin, D. J.; Trent, M. B.; Boor, P. J. Toxicology 1999, 138, 137-154. (188) McDonald, I. A.; Lacoste, J. M.; Bey, P.; Palfreyman, M. G.; Zreika, M. J. Med. Chem. 1985, 28, 186-193. (189) Silverman, R. B.; Hiebert, C. K.; Vazquez, M. L. J. Biol. Chem. 1985, 260, 14648-14652. (190) Palfreyman, M. G.; McDonald, I. A.; Zreika, M.; Cremer, G.; Haegele, K. D.; Bey, P. J. Neural Transm. Suppl. 1993, 40, 101-111. (191) Palfreyman, M. G.; McDonald, I. A.; Bey, P.; Danzin, C.; Zreika, M.; Cremer, G. J. Neural Transm. Suppl. 1994, 41, 407-414. (192) Abbenante, G.; Prager, R. H. Aust. J. Chem. 1992, 45, 1791-1800. (193) Abbenante, G.; Hughes, R.; Prager, R. H. Aust. J. Chem. 1994, 47, 1441-1452. (194) Dieter, R. K.; Li, S.-J. J. Org. Chem. 1997, 62, 7726-7735. (195) Hartwig, J. F. Angew. Chem., Int. Ed. Eng. 1998, 37, 2046-2067. (196) Arcadi, A.; Bernocchi, E.; Cacchi, S.; Caglioti, L.; Marinelli, F. Tetrahedron Lett. 1990, 31, 2463-2466. (197) McCarthy, J. R.; Matthews, D. P.; Edwards, M. L.; Stemerick, D. M.; Jarvi, E. T. Tetrahedron Lett. 1990, 31, 5449-5452. (198) McCarthy, J. R.; Huber, E. W.; Le, T.-B.; Laskovics, F. M.; Matthews, D. P. Tetrahedron 1996, 52, 45-58. (199) McCarthy, J. R.; Peet, N. P.; LeTourneau, M. E.; Inbasekaran, M. J. Am. Chem. Soc. 1985, 107, 735-737. (200) McCarthy, J. R.; Matthews, D. P.; Paolini, J. P. In Org. Synth., Coll. 1998; Vol. IX, p 446-449. (201) McCarthy, J. R.; Matthews, D. P.; Paolini, J. P. In Org. Synth., Coll. 1998; Vol. IX, p 442-446. (202) Harpp, D. N.; Gingras, M. J. Am. Chem. Soc. 1988, 110, 7737-7745. (203) Chen, C.; Wilcoxen, K.; Zhu, Y.-F.; Kim, K.-i.; McCarthy, J. R. J. Org. Chem. 1999, 64, 3476-3482. (204) Chen, C.; Wilcoxen, K.; Huang, C. Q.; Strack, N.; McCarthy, J. R. J. Fluorine Chem. 2000, 101, 285-290. (205) DeBoos, G. A.; Fullbrook, J. J.; Owton, W. M.; Percy, J. M.; Thomas, A. C. Synlett 2000, 963-966. (206) Chen, C.; Wilcoxen, K.; Strack, N.; McCarthy, J. R. Tetrahedron Lett. 1999, 40, 827-830. (207) Hadida, S.; Kim, S.-Y.; Curran, D. P. Unpublished results, University of Pittsburgh. (208) Larhed, M.; Hoshino, M.; Hadida, S.; Curran, D. P.; Hallberg, A. J. Org. Chem. 1997, 62, 5583-5587. (209) Genêt, J. P.; Blart, E.; Savignac, M. Synlett 1992, 715-717. (210) Blart, E.; Genêt, J. P.; Safi, M.; Savignac, M.; Sinou, D. Tetrahedron 1994, 50, 505-514. (211) Ramaiah, M. Tetrahedron 1987, 43, 3541-3676. (212) Maleczka Jr., R. E.; Lavis, J. M.; Clark, D. H.; Gallagher, W. P. Org. Lett. 2000, 2, 3655-3658.

73 10 Acknowledgements

This investigation was carried out at the Institution of Organic Pharmaceutical Chemistry, Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Uppsala.

The author would like to express his sincere gratitude to the following:

My supervisor, professor Anders Hallberg for invaluable help and for surrounding himself and the department with empathy and positive energy.

Hà for everything…

Elsa and Per-Ivan for neverending support and for letting me do things my way.

My four new sisters, their families and my mother-in-law for spring-rolls and sympathy.

Dr Mats Larhed for help and assistance during the writing of our papers and the thesis, Sun-Young Kim for help with the synthesis of the fluorous starting materials and Helena Sahlin for help with the synthetic work.

My colleagues at the department, especially my room mates Susanna Lindman, Ulrika Rosenström and Karl Vallin.

Mathias Alterman, Professor Anders Hallberg, Docent Anette Johansson, Dr Mats Larhed, Peter Nilsson, Ulrika Rosenström and Karl Vallin for proof-reading and constructive criticism.

Arne Andersson, Björn Carlsson and Marianne Åström for assistance whenever needed and Lotta Wahlberg for kind help with the chemicals.

George Paul John and Ringo, Monty Python, Norrbotten, Liljeborns Trafikskola.