In Copyright - Non-Commercial Use Permitted Rights / License: … · 2020. 3. 26. · List...

191
Research Collection Doctoral Thesis Ruthenium/PNNP-catalyzed asymmetric Michael addition Author(s): Santoro, Francesco Publication Date: 2007 Permanent Link: https://doi.org/10.3929/ethz-a-005361343 Rights / License: In Copyright - Non-Commercial Use Permitted This page was generated automatically upon download from the ETH Zurich Research Collection . For more information please consult the Terms of use . ETH Library

Transcript of In Copyright - Non-Commercial Use Permitted Rights / License: … · 2020. 3. 26. · List...

  • Research Collection

    Doctoral Thesis

    Ruthenium/PNNP-catalyzed asymmetric Michael addition

    Author(s): Santoro, Francesco

    Publication Date: 2007

    Permanent Link: https://doi.org/10.3929/ethz-a-005361343

    Rights / License: In Copyright - Non-Commercial Use Permitted

    This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.

    ETH Library

    https://doi.org/10.3929/ethz-a-005361343http://rightsstatements.org/page/InC-NC/1.0/https://www.research-collection.ethz.chhttps://www.research-collection.ethz.ch/terms-of-use

  • Diss. ETHNo. 17024

    Ruthenium/PNNP-Catalyzed Asymmetric

    Michael Addition

    A dissertation submitted to the

    SWISS FEDERAL INSTITUTE OF TECHNOLOGY ZURICH

    For the degree of

    DOCTOR OF SCIENCE ETH

    Presented by

    Francesco Santoro

    Dipl. Chem. ETH Zurich

    Born on July 15, 1979

    Accepted on Recommendation of

    Prof. Dr. Antonio Togni, Examiner

    Dr. Antonio Mezzetti, Co-Examiner

    Prof. Dr. Peter Chen, Co-Examiner

    Zurich 2007

  • Acknowledgements

    I would like to express my gratitude to Prof. Dr. Antonio Togni for giving me the privilege

    to carry out my Ph.D. in his group, for always being available whenever I had questions, and for the

    interesting discussions about chemistry and other topics.

    A special thanks to Dr. Antonio Mezzetti, who supervised my Ph.D. project, and

    significantly contributed to my scientific formation. His critic and zealous direction of my work

    stimulated, through numerous events, the continuous need for personal and professional

    improvements.

    I wish to thank Prof. Dr. Peter Chen for accepting to co-referee this thesis.

    I thank all the people who contributed to the development of this thesis through their

    technical expertise: Dr. Heinz Ruegger and Serena Filippuzzi for the NMR spectroscopy; Dr.

    Isabelle Haller, Dr. Sebastian Gischig, and Pietro Butti for the X-ray crystallography.

    I wish to thank all the members of the Togni group for the stimulating and pleasant working

    environment, and for the many lovely activities inside and outside the laboratory (including:

    dinners, holidays, conferences, ski weekends, weddings, concerts, cheese-and-wine(-and-movie)

    evenings, expovina excursions, dancing nights, all-night-up-and-cheeseburger-at-5-o'clock nights,

    etc.). In particular, I wish to thank my lab-mates, Patrick Eisenberger and Dr. Céline Réthoré, for

    the infinite patience and friendship shown every day. A big thanks to Pietro Butti and Marco

    Ranocchiari for the exceptional friendship we developed, which is very important to me.

    I would like to thank all my friends outside the group for their support and sincere friendship

    during my stay in Zürich. Among them, I have to express my gratitude to Luca Cereghetti, who had

    the guts to live under the same roof with me for 9 years (!!!!), and to Marco Abbondio, a true friend

    since childhood and very talented artist, who keeps inviting me to exquisite home-made dinners

    knowing that I could never keep up with his cooking standards.

    Infine vorrei ringraziare di cuore i miei genitori, mia sorella e Isabelle per avermi sostenuto,

    incoraggiato e amato incondizionatamente anche - e soprattutto - durante i periodi piu bui.

  • List of Publications

    "Chiral Dicationic Bis(aqua) Complexes [Ru(OH2)2(PNNP)]2 : The Effect of Double Chloride

    Abstraction on Asymmetric Cyclopropanation", C. Bonaccorsi, F. Santoro, S. Gischig, A. Mezzetti,

    Organometallics 2006, 25 (8), 2002-2010.

    "Chiral Ruthenium PNNP Complexes of Non-Enolized 1,3-Dicarbonyl Compounds: Acidity and

    Involvement in Asymmetric Michael Addition", M. Althaus, C. Bonaccorsi, A. Mezzetti, F.

    Santoro, Organometallics, 2006, 25 (13), 3108-3110.

    "Chiral Ruthenium PNNP Complexes in Asymmetric Michael Addition", F. Santoro, M. Althaus,

    C. Bonaccorsi, A. Mezzetti; in preparation.

  • 1 Table of Contents

    Table of Contents

    Abstract iii

    Zusammenfassung vRiassunto vii

    1 Introduction 1

    1.1 Lewis Acids in Organic Synthesis 1

    1.1.1 The Strength of Lewis Acids 3

    1.1.2 Carbonyl Groups Coordination to Lewis Acid 6

    1.1.3 C-H Acidity Enhancement by Lewis Acids Coordination 8

    1.1.4 Lewis Acids Coordination to Enones - Conjugate Addition 9

    1.2 Michael Addition 15

    1.2.1 Discovery and Definition 15

    1.2.2 Asymmetric Michael Addition 17

    1.2.3 Asymmetric Catalytic Michael Addition 18

    1.3 Heterofunctionalization of 1,3-Dicarbonyl Compounds 34

    1.4 Ruthenium/PNNP-Catalyzed Organic Transformations^^_

    39

    1.4.1 Chiral Tetradentate Ligands 39

    1.4.2 Epoxidation of Olefins 42

    1.4.3 Cyclopropanation of Olefins_______

    44

    1.4.4 Fluorinationof ß-Keto Esters 47

    1.4.5 Hydroxylation of ß-Keto Esters and oc-Acyl Lactams 50

    1.5 Aims__

    51

    1.6 Literature____^^^_^^^__„__„_„„__„

    54

    2 Ruthenium/PNNP-Catalyzed Michael Addition 612.1 Ruthenium/PNNP Complexes of 1,3-Dicarbonyl Compounds

    _______________

    61

    2.1.1 Complexes of Non-Enolized 1,3-Dicarbonyl Compounds 61

    2.1.2 Enolato Complexes 622.1.3 Note on the Rarity of Complexes With Non-Enolized 1,3-Dicarbonyl Compounds 65

    2.2 Catalytic Michael Addition 68

    2.2.1 Discussion on the Substrate Screening 70

    2.3 Activation Screening 73

    2.4 Stoichiometric Reactions 77

    2.4.1 Reaction of 3 with Methyl Vinyl Ketone.____m_____________^_

    77

    2.4.2 Reaction of 4 with Methyl Vinyl Ketone 782.5 Enolato Complexes as Catalyst Precursors 80

    2.6 The Effects of Polar Additives as Proton-Transfer Agents^_^^__„_„„

    82

    2.7 Solvent Effect______^^_

    83

    2.8 NMR Investigations of the Reaction of [RuCl2(la)] with (Et30)PF6 (2 equiv) 85

    2.9 pKa Determination of 1,3-Dicarbonyl Complexes 3 and 17 89

    2.9.1 Method^_^_____

    90

    2.9.2 Involvement in the Michael Addition 93

    2.10 Deuteration and Protonation of Enolate Complex 4 at Low Temperature 93

    2.10.1 Method 95

    2.10.2 Deuteration with CF3C02D....

    95

  • Table of Contents ii

    2.10.1i Deuteration with D2S04 96

    2.10.' Protonation with HBF^EtîO 98

    2.11 Concluding Remarks 100

    2.12 Literature 101

    3 Developing New Lewis Acids Based on Nitrosyl Complexes3.1 Introduction

    103

    103

    3.1.1 Counting Electrons 104

    3.1.2

    3.1.3

    3.1.4

    Frontier Molecular Orbitals Interpretation of Nitrosyl-Metal complexes

    Nitrosyl Complexes of Transition-Metals as Lewis Acids

    Metal Nitrosyl Complexes as Electron-Switch

    105

    108

    110

    3.2 Synthesis of rRuCl(PNNP)(NO)l 112

    3.2.1

    3.2.2

    3.2.3

    Synthesis of [Ru(dppe)2(NO)l+ (97a)Attempted Synthesis of [RuCl(la)(NO)l (100)

    Synthesis of [RuCl(lb)(NO)] (7) and Attempted Synthesis of [Ru(lb)(NO)]+

    112

    114

    115

    3.2.4

    3.2.5

    3.2.6

    Crystal Structure of rRuCl(lb)(NO)l (7)

    Crystal Structure of rRuCl(N02)(lb)l (8)

    Catalytic Attempts

    117

    122

    124

    3.3 Rh/PNNP Complexes 127

    3.3.1

    3.3.2

    3.3.3

    3.3.4

    Attempted Synthesis of [Rh(PNNP)(NO)l2+ (113)Crystal Structure of rRh(la)lPFrt (9)

    Synthesis of [Rhi2(la)l+ (10) and Catalytic AttemptsSynthesis of rRhMeI(la)l+ (11)

    128

    129

    130

    133

    3.3.5

    3.4

    Reaction of [Rh(la)]+ with (Me30)BF4: Synthesis and Characterization of 12Conclusion and Outlook

    136

    139

    3.5 Literature 141

    4 Experimental Part 143

    4.1 General 143

    4.2 Chapter 2 145

    4.2.1 Complexes 145

    4.2.2 Substrates 150

    4.2.3 Catalysis Products 154

    4.2.4 Activation Screening 157

    4.2.5 Stoichiometric Reactions 158

    4.2.6

    4.2.7

    4.3

    Measurement of the pKa of Complex 17

    Deuteration and Protonation at Low Temperature

    Chapter 3

    159

    160

    160

    4.4 Literature 166

    5 Appendix 167

    5.1 List of Abbreviations 167

    5.2 List of Substances 168

    5.3 Crystallographic Data 169

    5.3.1 rRuCl(lb)(NO)l (7) 169

    5.3.2 [RuCl(N02)(lb)l (8) 172

    5.3.3 |Rli(la)lPFfi (9PF6) 175

    5.4 Curriculum Vitae 179

  • iii Abstract

    Abstract

    The first part of this thesis (chapter 2) is focused on the application of the ruthenium

    complex [RuCl2(la)] (2) bearing a chiral tetradentate PNNP ligand (la) as catalyst precursor in the

    asymmetric Michael addition of 1,3-dicarbonyl compounds to methyl vinyl ketone.

    Previous studies revealed that the activation of [RuCl2(la)] (2) with halide scavengers gives

    a Lewis acidic catalyst that has been employed in the asymmetric fluorination and hydroxylation of

    1,3-dicarbonyl compounds. Differently than analogous systems, the ß-keto ester 5a coordinates to

    the activated ruthenium fragment in the ^o/7-enolized form to give 3. The structure of 3 was

    disclosed by comparing its NMR spectroscopic data with those of its enolato analogue 4, which was

    structurally characterized. Complex 3 catalyzes the Michael addition of 1,3-dicarbonyl compounds

    to methyl vinyl ketone in CH2C12 with quantitative yield and up to 93 % ee.

    QN

    ^-^Ph2 Ph2x=/1a

    0 O

    | 1) (Et30)PF6 (2 equiv)

    N7, l~N^

    ^P | P^CI&x ®&

    ~\2+

    NEt,

    OFT

    2 + 2(Et30)PF6

    (5 mol-%)

    R2

    5

    up to >99 % yield, 93 % ee

    Stoichiometric and catalytic experiments show that the enolato adduct 4 is not nucleophilic

    enough to react with the enone. The p/Ca of two 1,3-dicarbonyl-ruthenium adducts was measured

    and, although the acidity of the p-keto ester is enhanced by about 8 orders of magnitude, it is still at

    least 6 orders of magnitude lower than that of methyl vinyl ketone. This strongly suggests that the

    enone is not protonated to a significant extent in the reaction mixture. Taken together, these data

    suggest that complex 3, rather than 4, is an intermediate in the catalytic reaction.

    On the other hand, the catalytic performances are influenced by the presence of a polar co-

    solvent such as Et20. Low temperature deuteration and protonation of the enolato complex 4

    suggest the existence of a thermally disfavored tautomer of 3, that we tentatively identify as the

    coordinated enol compound 3'.

  • Abstract iv

    3 n 3.

    A preliminary reaction mechanism is proposed in which 3' is the active species in the C-C

    bond formation step that activates and attacks the enone in a single step. Thus, Et20 is proposed to

    act either as proton shuttle to assist the deprotonation of 3, or to favor the tautomerization of 3 to 3'.

    The fact that the speed and enantioselectivity enhancement of the catalysis are dependent by the

    kind of ether used supports this hypothesis.

    The second part of this thesis describes the efforts to prepare new Lewis acidic catalysts

    based on tetradentate PNNP ligands for the asymmetric cyclopropanation and epoxidation of

    olefins. In this context, two strategies have been undertaken.

    Firstly, it has been attempted to use the bending of the nitrosyl in {MNO}8 complexes to

    promote the transfer of a carbene or an oxo ligand from a reactive source to the metal, and then to

    the substrate. To that end, a nitrosyl complex of formally ruthenium(O) (7) was prepared and

    structurally characterized. To the best of our knowledge, 7 is the first {RuNO}8 compound with

    octahedral geometry and bent mctal-nitrosyl bond. Complex 7 reacts readily with molecular oxygen

    to give the nitro complex 8, which was also structurally characterized.

    As second strategy, we prepared new rhodium-PNNP complexes to be used as Lewis acidic

    catalysts, in analogy with the ruthenium-PNNP chemistry, but with the advantage of the double

    positive charge. To that end, complexes [RhI2(la)]PF6 (10), [RhIMe(la)]PF6 (11), and

    [RhMe(la)](PF6)2 (12) were prepared. Moreover, the structure of the precursor 9 was determined.

    Q

    1b

    /IN',, ' ,,'INn

    CI 7

    hJLh

    \* I VCI 8

    /=N N-

    \Jriph2 Ph2

    1a

    cN>op pP'

    9

    n +

  • V Zusammenfassung

    Zusammenfassung

    Der erste Teil dieser Dissertation (Kapitel 2) umfasst die Anwendung des

    Ruthenium-Komplexes [RuCi2(la)] (2) als Katalysator für die asymmetrische Michael Addition

    von 1,3-Dicarbonylverbindungen an Methylvinylketon.

    Wie bereits berichtet, führt die Aktivierung von [RuCbCla)] (2) mit Chloridabstraktoren zur

    Bildung eines Lewis-aziden Katalysators, der für die asymmetrische Fluorierung und

    Hydroxyherung von 1,3-Dicarbonylverbindungen angewendet wurde. Im Gegensatz zu analogen

    Systemen koordiniert ß-Ketoester 5a in seiner nicht-enolisierten Form ans Ru/PNNP-Fragment (3).

    Die Charakterisierung von Komplex 3 erfolgte mittels NMR-Spektroskopie und des Vergleiches

    mit den NMR-Daten des analogen Enolatokomplexes 4, dessen Struktur kristallographisch

    bestimmt wurde. Komplex 3 katalysiert die Michael Addition von 1,3-Dicarbonylverbindungen an

    Methylvinylketon in CH2CI2 mit quantitativen Ausbeuten und Enantioselektivitäten bis 93 % ee.

    QN=

    \}~p p

    ^-yPh2 Ph2

    1a

    O O

    CI 1)(Et30)PF6(2equiv

    cfe>2> °H2+

    CI

    2

    OR^

    R2

    5

    2 + 2(Et30)PF6

    (5 mol"%bis zu >99 % Ausbeute

    93 % ee

    Stöchiometrische und katalytische Experimente zeigen, dass der Enolatokomplex 4 nicht mit

    Methylvinylketon reagiert, weil er nicht nucleophil genug ist. Der pKa von 3 wurde gemessen und

    zeigt, dass die Azidität des a-protons des ß-Ketoesters in 3 um acht Grössenordnungen größer ist

    als in freiem 5a. Trotzdem, Methylvinylketon ist immer noch sechs Grössenordnungen saurer als 3.

    Diese Vergleiche legen nahe, dass Methylvinylketon unter den Reaktionsbedingungen nicht

    protoniert wird. Es zeigt sich daher, dass Komplex 3, und nicht 4, ein Zwischenprodukt in der

    katalytischen Reaktion ist.

    Anderseits beeinflusst die Anwesenheit von polaren Lösungsmitteln wie Et20 die

    katalytischen Eigenschaften des beschriebenen Systems. Tieftemperatur-Deuterierungen und

    Protonierungen des Enolatokomplexes 4 weisen auf die Existenz eines thermodynamisch

    unbevorzugten Tautomers von 3 hin, das wir als den Enol-Komplex 3' formuliert haben.

  • Zusammenfassung vi

    N^\~|2+-h-tQ

    3 n 3'

    Ein Reaktionsmechanismus wurde vorgeschlagen, in dem 3' die reaktive Spezies in der

    Bildung der C~C Bindung ist, wobei das Enon gleichzeitig durch Wasserstoffbrücken aktiviert und

    vom Nucleophil angegriffen wird. Es wird daher vorgeschlagen, dass Diethylether entweder als

    Protonentransfer-Reagenz wirkt, oder die Tautomerisierung von 3 zu 3' fördert. Die Abhängigkeit

    der Katalyseeigenschaften von der Art des Esthers unterstützt diese Hypothese.

    Der zweite Teil dieser Dissertation befasst sich mit dem Versuch, neue Lewis-azide

    Katalysatoren, basierend auf PNNP Liganden, für die asymmetrische Cyclopropanierung und

    Epoxidierung zu synthetisieren. Zwei verschiedene Strategien wurden dabei verfolgt.

    Erstens wurde versucht, die Koordinationseigenschaften des Nitrosylliganden in {MNO}

    Komplexen anzuwenden, um die Übertragung eines Carben- oder Oxo-Liganden aus einer

    entsprechenden reaktiven Quelle zum Metall und dann zum Substrat zu fördern. Ein formal

    Ruthenium(0)-Nitrosylkomplex (7) wurde synthetisiert und dessen Kristallstruktur bestimmt. Nach

    unserem bestem Wissen ist 7 das erste Beispiel eines {RuNO}8 Komplexes mit oktaedrischer

    Geometrie und gewinkelter Metall-Nitrosyl Bindung. Komplex 7 reagiert mit molekularem

    Sauerstoff zum Mitrit-Komplex 8, dessen Struktur kristallographisch untersucht wurde.

    Der zweite Ansatz befasst sich mit der Synthese von Rhodium/PNNP-Komplexen als

    Analoge zu den Ruthenium/PNNP Komplexen, aber mit dem Vorteil, dass sie zweifach geladen

    sind. Die Komplexe [RhI2(la)]PF6 (10), [RhIMe(la)]PF6 (11) und [RhMe(la)](PF6)2 (12) wurden

    synthetisiert und die Struktur von Komplex 9 wurde kristallographisch bestimmt.

    \}~p p"\ /^-?

    Ph2 Ph2x=/

    1b

    nCT^Ns

    c.\ -,

    N

    p

  • Vil Riassunto

    Riassunto

    La prima parte di questa tesi (capitolo 2) è incentrata sull'applicazione del complesso di

    rutenio [RuChfla)] (2) contenente un legante tetradentato chirale PNNP (la) quale precursore del

    catalizzatore nell'addizione di Michael asimmetrica di composti 1,3-dicarbonilici al metil¬

    vinilchetone.

    E' stato osservato in precedenti studi che l'estrazione di un legante cloro da [RuC^la)] (2)

    produce degli acidi di Lewis che sono stati utilizzati corne catalizzatori nella fluorurazione e

    idrossilazione asimmetrica di composti 1,3-dicarbonilici. A differenza di sistemi catalitici analoghi,

    il ß-chetoestere 5a si coordina al complesso attivo di rutenio nella forma non enolizzata per dare il

    complesso 3. La struttura di 3 è stata dedotta confrontando i dati di spettroscopia NMR con quelli

    dell'analogo enolato complesso 4, per il quale è stato possibile determinare la struttura cristallina. Il

    complesso 3 catalizza l'addizione di Michael di composti 1,3-dicarbonilici al metilvinilchetone con

    rese quantitative e fino al 93% di eccesso enantiomerico (ee).

    Q/=N N=n

    1a

    1)(Et30)PF6(2equiv)

    "db^,,' oy, y*

    ClV~y

    5a

    0 0O

    2 + 2(Et30)PF6

    OR3+ ^JL (5 m°' %) , R1 ,/\ ORJ fino a >99 % di resa, 93 % ee

    R2

    5

    Esperimenti stechiometrici e catalitici hanno dimostrato che l'enolato complesso 4 non è

    abbastanza nucleofilo per reagire con l'enone. È stato misurato il pK^ di due addotti di rutenio con i

    substrati 5a e 5g e, per quanto l'acidità del substrata sia stata incrementata di otto ordini di

    grandezza, è sempre minore dell'acidità dell'acido coniugato dell'enone di almeno un fattore 6.

    Questo suggerisce che l'enone non venga significativamente protonato durante la catalisi. In

    conclusions queste osservazioni suggeriscono che 3 (e non 4) sia un intermedio del ciclo catalitico.

    È stato osservato che la reazione catalitica è influenzata dalla presenza di solventi polari

    quali per esempio Et20. Analisi NMR délia reazione di deuterazione e protonazione dell'enolato

    complesso 4 a bassa temperatura suggeriscono la presenza di un tautomero di 3

    termodinamicamente sfavorito. Si presuppone che questo composta sia l'enolo complesso 3'.

  • Riassunto vin

    3 n 3.

    Si propone quindi, in via präliminare, un meccanismo di reazione in cui 3' è la specie attiva

    nella la tormazione del nuovo legame carbonio-carbonio, che al tempo stesso attiva e attacca

    1'enone. Di conseguenza si è ipotizzato che l'etere assista la prototropia: promuovendo la

    deprotonazione di 3 0 favorendo la tautomerizzazione da 3 a 3'. Questa ipotesi è supportata dal fatto

    che la velocità e la stereoselettività della reazione catalitica dipendano dal tipo di etere usato.

    La seconda parte di questa tesi descrive i tentativi di sviluppo di nuovi acidi di Lewis basati

    su complessi di leganti tetradentati PNNP quali catalizzatori asimmetrici per la ciclopropanazione e

    l'cpossidazione di olefine. Questo obiettivo è stato perseguito attraverso due stratégie.

    Inizialmente si è cercato di sfruttare le modalité di coordinazione del nitrosile nei complessi

    del tipo {MNO}8 per promuovere il trasferimento di un carbene 0 ossene da reagenti reattivi al

    métallo, e quindi al substrato. A questo scopo è stato preparato un complesso nitrosilico di rutenio

    (7), del quale è stata determinata la struttura crystallina. Per quanto ci risulta, 7 è il primo composto

    del tipo {RuNO}8 ad avere una geometria ottaedrica e un legame M-N-0 piegato. II complesso 7

    reagisce velocemente con ossigeno molecolare producendo il nitro complesso 8, la cui struttura

    cristallina è stata ugualmente determinata.

    Come seconda strategia sono stati preparati dei nuovi complessi di rodio con leganti PNNP

    da usare come catalizzatori in maniera analoga ai succitati complessi di rutenio, awalendosi

    dell'incremento di acidità di Lewis dovuto alla doppia carica positiva. A questo scopo sono stati

    sintetizzati i complessi [RhI2(la)JPF6 (10), [RhIMe(la)]PF6 (11) e [RhMe(la)](PF6)3 (12). La

    struttura cristallina del complesso 9 è stata determinata.

    Qtfi

    ,

    £&Ph2 Ph2W1b

    CI 7

    h^Lh-N.„ I „>N>

    Cp-^P)CI 8

    Qn +

    1

    tfg=N N=

    9

  • 1 Introduction 1

    1 Introduction

    This thesis is focused on the development of an asymmetric Michael addition catalyzed by

    Lewis-acidic ruthenium/PNNP complexes. Therefore, the introduction will cover the relevant

    aspects of Lewis acids in organic chemistry, an overview on the Michael addition with focus on its

    asymmetric version, and the application of the ruthenium/PNNP catalysts developed in our

    laboratory.

    1.1 Lewis Acids in Organic Synthesis

    G. N. Lewis published in 1923 a landmark essay laying the foundation for our modern

    understanding of the valence bond.1 In this work, he formulated the concept of the chemical bond

    by the sharing of electron pairs and provided the following definition of donors and acceptors:

    "...with complete generality we may say that a basic substance is one which has a lone pair of

    electrons which may be used to complete the stable group of another atom, and that an acidic

    substance is one which can employ a lone pair from another molecule in completing the stable

    group of one of its own atoms." In modern terms, a Lewis acid is a chemical entity with an empty

    orbital that is able to accept an electron-pair from the filled orbital of another molecule—the base.

    Although the concept of Lewis acidity is different from the definition of Bransted acidity, it must be

    noted that there is no such difference between Lewis bases and Bronsted bases. Both concepts are in

    fact equivalent because all species able to donate an electron-pair are, to a certain extent, proton

    acceptors.

    The involvement of Lewis acids in organic chemistry is associated with their complexation

    with organic molecules, usually, at electron-rich heteroatomic functional groups. Among these, the

    interaction between Lewis acids and carbonyl groups received the greatest attention from the

    scientific community because of its central role in synthetic organic chemistry. The importance of

    Lewis acids in synthetic organic chemistry is indicated by the tremendous amount of publications

    available in books,3"6 reviews, and articles. Scientific accounts concerning Lewis acids appear in

    virtually every edition of major chemistry journals.

    Initially, the Lewis acids applied for the promotion of Diels-Alder reactions were mainly

    halides of B(I1I), Sn(IV), Ti(IV), lanthanides, Zn(II), and Mg(II).7 At that time, stoichiometric or

    *

    In 1988 the International Union of Pure and Applied Chemistry (IUPAC) suggested to use the term "proton"

    only in relation to the cation of the isotope 'H and to use the word "hydron" in the case of isotope mixtures.2 However,

    the term "proton" is still by far the most commonly used expression to indicate the naturally occurring cation. For that

    reason, only the word "proton" will be used in this thesis.

  • 2 1 Introduction

    greater amounts of the Lewis-acidic reagent were employed to compensate for the decomposition

    by hydrolysis and, thus, to achieve the highest acceleration.

    However, the interest in Lewis-acidic compounds is mainly owed to their application as

    molecular catalysts. Lewis acids can catalyze an enormous number of different reactions, thus, any

    serious attempt to comprehensively survey all aspects of this chemistry is a frightful task. However,

    one of the most important contributions of Lewis acids in synthetic chemistry is their application in

    the formation of new carbon-carbon bonds.

    There is no need to point out the importance of the formation of new carbon-carbon bonds in

    organic synthesis. Classical reactions such as the Friedel-Crafts alkylation and acylation, Diels-

    Alder cyclization, ene reaction, Mukaiyama aldol reaction, epoxide opening, and conjugate

    additions are only a selection of the most important transformations promoted by Lewis acids such

    as BF3, AlCb, TiCl4, SnCl4, FeCl3 (Scheme 1.1).3"6

    R'

    o

    XR2 ;

    + R2-X

    OR

    Friedel-Crafts

    Acylation

    Friedel-Crafts

    Alkylation

    Die Is-Alder

    Reaction

    O^OR

    X = Halogen

    X = Halogen

    ROH

    + J^ ene ReactionOH

    C02R

    OTMS O

    + AH R

    Mukaiyama

    aldol Reaction AX

    Scheme 1.1: Examples of key organic transformations catalyzed by Lewis acids

    Unfortunately, these "classical" Lewis acids often do not provide the chemo-, regio-, or

    stereoselectivity required by modern organic synthesis. For that reason, the research in this field has

    been characterized in the last years by an impressive expansion in the development of more

  • 1 Introduction 3

    selective, reactive, and versatile catalysts. To that end, a Lewis acid can be coordinated to

    specifically designed ligands allowing the fine tuning of the reagent properties in terms of acidity

    strength, chemo-, stereo-, and, by means of chiral ligands, enantioselectivity. An exhaustive

    discussion on the role of Lewis acids in organic synthesis can be found in the books of Schinzer,

    and Santelli and Pons.4

    In the next paragraphs, it will be introduced how a Lewis acid can activate a substrate to

    promote reactions. To that end, the discussion will concern the concept of Lewis acidity and some

    attempts made to measure it. In a second part, it will be shown how Lewis acids can activate

    carbonyl compounds in three model reactions—the 1,2-addition to carbonyl groups, the a-

    functionalization of carbonyl groups, and the conjugate addition. In each case, the function of the

    Lewis acid will be highlighted by focusing on the interactions between the Lewis acid and the

    organic substrates.

    1.1.1 The Strength of Lewis Acids

    After the discovery by Yates and Eaton8 in 1960 that Lewis acids accelerate Diels-Alder

    reactions, chemists were interested in giving a quantitative classification of Lewis acids according

    to their strength. The idea of putting Lewis-acidic compounds in tables ordered by their acidity

    strength, as it has been done for Bransted acids, was an appealing task. Unfortunately, the strength

    of a Lewis acid is a more complicated concept than the strength of a Bransted acid.

    For instance, the latter is defined as the extent of proton dissociation in a solvent of

    reference. Bronsted acids can thus be ordered according to their strength by comparing the

    equilibrium constant of the proton transfer reaction reported in equation 1.1 with B being the

    solvent molecule.

    AH + B - A" + BH+ (Eq. 1.1)

    However, according to Lewis's concept of acids and bases, equation 1.1 is a particular case

    of the general base-exchange reaction formulated in equation 1.2 in which the only Lewis acid is

    H+. Thus, the comparison of Bronsted acidity means to compare the proton affinity of different

    compounds.

    AB1 + B2 - B1 + AB2(Eq 1,2)

    In the general case when the acid is not H+ but any Lewis acid, the equilibrium constant /Ceq

    in equation 1.2 is determined the by difference in the standard free energies of adducts AB1 and

    AB2.

  • 4 1 Introduction

    Measure ofStandard Gibb 's Free Energy

    Satchell and Satchell reported in 1969 a first attempt of quantitative measurement of Lewis

    acidity according to the equilibrium constants Keq of the association of a Lewis acid with a series of

    different bases in solution.9 This might still be described by equation 1.2 in which the base B1 is the

    solvent. The measure of equilibrium constants is inherently a measure of the standard Gibb's free

    energy AG0 according to equation 1.3.

    AG° = -RT\n(Keq) (Eq. 1.3)

    A major drawback of this method is the competing association of the solvent to the Lewis

    acid. In fact for solubility reasons, it was not always possible to use non-coordinating solvents such

    as benzene or hexane, and polar solvents such as acetone or diethyl ether had to be used. The

    authors fairly argued that within a series of experiments—that is, same Lewis acid, same solvent,

    different bases—all solvent effects were expected to be much the same, also because interactions of

    the solvent with the base are usually negligible.

    This method offers the possibility to compare the affinity of a Lewis acid for different bases

    but no direct comparison between Lewis acids is meaningful because each acid interacted

    differently with the solvent. Moreover, the method was restricted to 1:1 adducts.

    For that reason, this approach failed to give a satisfying quantitative scale of Lewis acidity

    strength.

    Measure ofStandard Enthalpy

    Drago et al. developed a method for the quantitative measurement of Lewis acidity based on

    the enthalpy difference AH° of the Lewis acid-base association reaction in the gas phase or in non-

    coordinating solvents (equation 1.4).10"13

    A + B AB

    (Eq. 1.4)-AH = EAEB + CACB-W

    The acid A and the base B were characterized by the empirical parameters E and C. The

    enthalpy of the reaction was partitioned into two terms EA EB and CA CB which were said to

    correspond to the electrostatic (ionic and dipole-dipole interaction) and covalent contribution in the

    acid-base interaction, respectively. The relative scales of parameters E and C for both the acid and

    the base have been fixed from four arbitrarily chosen reference values, and then statistically

    optimized from a large set of enthalpy values of adduct formations. A third parameter W has been

  • 1 Introduction 5

    introduced later to account for processes independent from the actual acid-base association reaction.

    For example in the reaction of the dimer [RhCl(CO)2]2 with a base B to give the adduct

    [RhCl(B)(CO)2], /^corresponds to the enthalpy of cleaving the dimer.

    In the case one or both reaction partners are charged, there is a significant stabilization by

    the partial charge-transfer between the acid and the base, which causes a poor fit of the

    experimental data. To account for this phenomenon, the third term W was replaced with RaTb

    where RA is the "receptance" of the acid and Tü the "transmittance" of the base (equation 1.5).14

    -AH = EAEB+ CA CB + RATB (Eq. 1.5)

    This model predicts satisfyingly the enthalpy of the association reaction between Lewis acids

    and bases, albeit only in the gas phase or very weakly-coordinating solvents such as CCU or

    benzene. The enthalpy of reactions carried out in more polar solvents can deviate substantially from

    the calculated value.13

    Anyway, this approach shows, basically, that the nature of the Lewis acid-base interaction is

    composed by electrostatic attraction, covalent bond resulting from the energy match and overlap

    between the filled donor orbital and the empty acceptor orbital, and charge-transfer stabilization

    energy.

    It must be noted that this model elegantly includes the concept of hard and soft acids and

    bases (HSAB),15'16 which is a very important feature governing Lewis acids-bases interactions. In

    fact, a hard acid or base will have a larger E- and a smaller C-value, whereas a soft acid or base will

    have a smaller E- and a larger C-value. In other words, interactions between hard acids and bases

    are prevalently of electrostatic nature, whereas interactions between soft acids and bases are

    essentially of covalent nature.

    Measure ofNMR Chemical Shifts

    In 1982, Childs17 offered a fundamentally original method for the determination of Lewis-

    acidic strength. He observed the chemical shifts differences in oc,ß-unsaturated carbonyl compounds

    upon coordination to Lewis acids. For example he observed that, for crotonaldehyde, the magnitude

    of the shift of the H signal upon complexation is the greatest, whereas those of H2 and H4 are

    smaller but still linearly related to the shift of the H3 signal. Instead, the shift of the H1 signal

    induced by coordination appeared to vary randomly (Figure 1.1).

  • 6 1 Introduction

    3

    44h-Y T 'HiH H

    4 2

    Figure 1.1

    Based on the shifts of the H3 resonances of various bases, the authors proposed the following

    scale of Lewis acidity: BBr3 > BC13 > SbCl5 > A1C13 > BF3 > EtAlCfe > TiCl4 > Et2AlCl > SnCL, >

    Et3Al.

    In 1990, Laszlo18 provided some theoretical support to the experimental work of Childs,

    Calculations of Lewis acid adducts of crotonaldehyde and methyl acrylate showed a direct

    correlation between the shifts of the H3 resonances with the LUMO (71*) energy of the oc,ß-

    unsaturated carbonyl substrate.

    The experimental confirmation of the reliability of such method was reported shortly after.

    The observed pseudo first-order rate constant k of the ene reaction catalyzed by Lewis acids showed

    a linear relation with the calculated energy of the LUMO orbital of methyl acrylate (Scheme 1.2).19

    Most importantly, the activity of the Lewis acids tested in this experiment showed the same trend as

    predicted by Childs.

    ^% -U, Lewis acid (10 mol-%)Scheme 1.2: Lewis acid-catalyzed ene reaction

    In conclusion, the NMR method developed by Childs has the considerable merit to provide a

    direct measurement of the strength of a Lewis acid expressed as activation power, which is most

    valuable in the prediction of catalytic activity.

    1.1.2 Carbonyl Groups Coordination to Lewis Acid

    Understanding the nature of the interactions between a Lewis acid and a carbonyl group is a

    necessary prerequisite to fully appreciate the function of Lewis acids in catalysis. Carbonyl

    compounds interact with Lewis acids primarily through a-coordination with one lone pair of the

    oxygen atom. Rare examples of ^-coordination through the C-0 double bond building a if-

    metalloxirane complex are also known but they are not relevant for the purpose of this discussion,7

  • 1 Introduction 7

    Representative examples of a- and 7t-bonded complexes of Lewis acid with carbonyl groups are

    illustrated in Figure 1.220,21

    ^-coordination

    Ri" R2

    X

    7i -coordination R2'/

    R

    M

    !X-b Ar'

    Figure 1.2: The a- and ^-coordination modes of Lewis acids with carbonyl groups

    The most important consequence of carbonyl coordination to a Lewis acid is the increase of

    electrophilicity at the carbonyl C-atom. The concept of enhanced electrophilicity of a carbonyl

    group upon Lewis acid coordination was explained by Reetz20 for the coordination of BF3 to

    acetaldehyde by means of semi-empirical MNDO calculations.

    H,C

    n* LUMO

    -0.776 °-585

    E = +0.76 eV

    -0.827

    E=-1.59eV

    BF,

    Figure 1.3: MNDO energies and coefficients of the LUMO of CH3CHO and CH3CHO-BF3

    The results showed that the BF3 complex had a lower LUMO energy and increased positive

    charge at the carbonyl C-atom (Figure 1.3 and Figure 1.4). As result of the altered electronic

    distribution, Lewis acids can promote the nucleophilic attack at the carbonyl C-atom and, thus,

    catalyze the 1,2-addition to carbonyl groups.

    H3C'S+0.242

    H3C H

    +0.363

    -BF,

    Figure 1.4: MNDO-computed charges of the carbonyl C-atom

  • 8 1 Introduction

    The addition to C-0 double bonds is a very important chemical transformation because of its

    very broad applicability. In fact, countless different carbon- and heteronucleophiles have been

    successfully added to carbonyl groups. Among the additions of carbon nucleophiles catalyzed by

    Lewis acids, the most important ones are the Mukaiyama aldol reaction, the reduction of the

    carbonyl group, and the alkylation, allylation, cyanation, and phenylation of aldehydes, ketones, and

    esters (Scheme 1.3).

    O Lewis acid OH

    Rl + r*ar3 Ar

    OSiMe3

    R1 = R^^ H" CN" AlkyP Ar"Mukaiyama reduction cyanation alkylation arylationaldol reaction

    Scheme 1.3: Examples of nucleophilic addition to carbonyl compounds catalyzed by Lewis

    acids

    1.1.3 C-H Acidity Enhancement by Lewis Acids Coordination

    The interaction of Lewis acids with non-conjugated carbonyl groups results in the increase

    of the acidity of the a-proton. This effect derives from the change in the charge distribution within

    the molecule and from the stabilization of the enolate (Scheme 1.4).

    n

    -H ?"r

    .—-

    ^yRH H H

    LA^ LA.O CT

    R ^ü J^R

    H H H

    Scheme 1.4: Effect of the coordination of a carbonyl to a Lewis acid on the adjacent a-

    proton

    An example of exploitation of this effect in organic synthesis is the aldol condensation via

    boron enolates.22 In this reaction, a ketone is smoothly converted to the corresponding enolato-

    boron complex by deprotonation with an amine such as N'Pr2Et. The enolato-boron complex reacts

    then with the aldehyde to give the expected aldol product with high stereo- and regioselectivity

    (Scheme 1.5).

  • 1 Introduction 9

    BR2(OS02CF3) BR,„Xu

    0 (lequiv) 0'2 R H 0 OH Q OH

    riA^r2iprNB R1^ Ri^Y^RS

    + Rl^Y^R32

    R2 R2 R2

    R' = Alkyl Z/E ratio up to 99-1 erythro/threo ratio up to 97:3

    Scheme 1.5: Boron-promoted aldol reaction

    aq»-aBy complexation with the boron reagent, the a-H acidity of the ketone (normally pATa'

    -19-20)23 is increased by at least 8-9 pK3 units, so that 'Pr2EtN (pKam = ~\ l)24 is sufficiently basic

    to carry out the deprotonation. However, the Lewis acid must be used in stoichiometric amounts to

    avoid self-condensation along with other side-reactions, which rules out the development of a

    catalytic version of the direct aldol reaction.

    Enhancement of acidity by complexation with Lewis acids is particularly evident with 1,3-

    dicarbonyl compounds. These compounds already contain an acidic C-H group (pKa = 10-15) and,

    upon coordination with a Lewis acid, they usually enolize spontaneously without the need of a

    supplementary base. The synthetic applications of the Lewis acid activation of 1,3-dicarbonyl

    compounds will be discussed in Paragraphs 1.2 and 1.3.

    1.1.4 Lewis Acids Coordination to Enones - Conjugate Addition

    A particular class of substrates that can be activated by coordination to Lewis acids are

    a,ß-unsaturated carbonyl compounds. This interaction promotes two classic organic

    transformations: the Diels-Alder reaction and the conjugate addition. For the purpose of this

    introduction, only the conjugate addition will be discussed here.

    The nucleophilic addition to a conjugated system is of paramount importance in organic

    synthesis because it is one of the oldest and most frequently used methods for the formation of new

    C-C bonds.25,26 Differently to classical 1,2-additions, the conjugate addition occurs at the

    extremities of the conjugated system (Scheme 1.6).

    1,2-additionX-Y - ^s^^O-

    X

    1,4-addition

    ^^° + X-Y - X^^/0.y

    Scheme 1.6: Simplified concept of 1,2-addition versus 1,4- or conjugate addition

  • 10 1 Introduction

    For example, the cyanide ion, which normally adds to the carbonyl bond in aldehydes and

    ketones, can add to a C-C double bond of a conjugated system (Scheme 1.7).25

    EtOH

    Ph^^^O + KCN » Ph^\^0Ph CN Ph

    Scheme 1.7: Conjugated addition of cyanide

    Although the reaction appears to be a 1,2-addition to the C-C double bond, the mechanism

    actually involves the 1,4-addition of HCN to the conjugated system followed by tautomerization of

    the enolate intermediate to the more stable keto form (Scheme 1.8).25

    Ph^^\^0 + HCN - Ph^^^OH __—^ Ph^/\^0Ph CN Ph CN Ph

    Scheme 1.8: Stepwise addition of HCN to the conjugated system showing the enol

    intermediate

    The electrophile is typically an a,ß-unsaturated carbonyl compound (ketone, aldehyde, or

    ester), but alkenes with other electron withdrawing groups such as nitro, cyano, or sulfonyl are also

    reactive electrophiles in the conjugate addition. Similarly, different kinds of nucleophiles such as

    amines, phosphines, silanes, alkoxides, and sulfides have also been used (Scheme 1.9).

    Nu^^^EWG 1) Nu~ l

    R^

    2) H+

    Nu = -C, -N,-P,-Si, -O, -S...

    EWG = -CHO, -COR, -C02R, -CONR2, ~N02, -CN, -S02R

    Scheme 1.9: Generalized conjugate addition

    The concept of conjugate addition can be extended to longer conjugated systems, as

    demonstrated in the examples reported in Scheme 1.10.27,2 However, it must be noted that these are

    rare reactions as compared to the classical 1,4-addition.

  • 1 Introduction 11

    <,C02Et 1,6-addition

    CQ2Et

    0 02S'^OH

    1,8-addition

    Scheme 1.10: Examples of 1,6- and 1,8-conjugate additions

    Frontier Molecular Orbital Considerations

    Most of the early studies on the coordination of Lewis acids to a,ß-unsaturated carbonyl

    compounds were performed in association with Diels-Alder reactions. However, some of these

    findings are also relevant to the conjugate addition.

    Lewis acids can promote the conjugate addition as indicated by the example reported in

    Scheme 1.11. Here, the addition of AlEt3 to the reaction mixture leads to a faster reaction (6 vs. 29

    hours) and to the smooth formation of the conjugate adduct (>99 % vs. 67 % yield) even when the

    ß-carbon of the enone is highly substituted.29

    HMethod A: HCN, NH4CI. 29 h

    Method B: HCN, AICI3, 6 h

    CN

    Method A: 67 % yield, 1:2 cis/trans ratio

    Method B: >99 % yield, 1:9 cis/trans ratio

    Scheme 1.11: Comparison of conjugate addition ofHCN with and without Lewis acid

    It must be considered that a,ß-unsaturated carbonyl compounds are ambidentate

    electrophiles because the nucleophilic attack can occur either at the ß-carbon or at the carbonyl

    carbon atom. Although the conjugate addition affords more stable products, the electron deficiency

    at the carbonyl carbon is greater than at the ß-carbon. Despite that, the conjugate addition is the

  • 12 1 Introduction

    most commonly observed process. The factors that favor the 1,2-regioselectivity as compared to the

    1,4-addition have been elucidated by Anh and co-workers according to the frontier molecular

    orbitals (FMO) theory.30"33 DFT calculations performed on acrolein as standard substrate revealed

    that the FMO coefficient of the LUMO is larger at the ß-carbon.34 For soft nucleophiles, such as

    delocalized carbanions, the frontier orbital interaction is more important than electrostatic effects;

    thus, the attack at the ß-carbon predominates. In other words, the conjugate addition is an orbital

    controlled process.

    Houk34 and Ottenbrite35 performed DFT calculations on acrolein as model substrate and

    compared the frontier molecular orbitals (FMO) of the free compound with its BF3-complex and the

    protonated species. The results showed that, upon coordination of a Lewis acid or protonation, the

    most noticeable change in the electronic structure of the aldehyde is the considerable decrease of

    the energy of the LUMO (Figure 1.5).

    .^\-^-0 j^-^'Z'O^ ^\^-0*

    E = +5.96eV E = +3.33eV E=-6.5eV

    Figure 1.5: FMO energies of acrolein, acrolein-BF3 complex, and acroleinium ion

    The reduced energy gap between the HOMO of the diene and the LUMO of the dienophile is

    believed to be responsible for the observed enhancement of the reaction rate.

    Conjugate Addition ofReactive Carbanions

    To form new C-C bonds, the nucleophile must be a carbanion. It is possible to divide

    carbanions in two classes. Stabilized carbanions are those in which the negative change is stabilized

    by derealization with a near electron withdrawing group such as ketones, aldehydes, esters, nitro

    groups, cyano groups, etc. The anions of relatively acidic C-H groups belong to this category. Non-

    stabilized carbanions are compounds in which the negative charge is localized. It is the case, for

    example, of simple alkyl or aryl organometallic reagents (Figure 1.6).

    AUS

    « , ... .

    i^\- 1^ 02N. _ NC. R102S\_Stabilized carbanions R] R^l ] | l

    R2 R2 RR R2

    Non-stabilized carbanions RLi ZnR2 RMgBr Li[R2Cu] AIR3

    Figure 1.6: Examples of stabilized and non-stabilized carbanions

  • 1 Introduction 13

    The conjugate addition of organometallic carbanions such as organolithium, organozinc,

    Grignard, and cuprate reagents to a,ß-unsaturated carbonyl compounds represents one of the most

    investigated reaction class and constitutes a conspicuous chapter in the literature of C-C bond

    formation. ' •- The conjugate addition with organometallic compounds is a strategic

    transformation from a synthetic point of view because it offers the possibility to form a new C-C

    bond with high chemo- and regioselectivity by using highly reactive reagents that are generally

    readily available from simple alkyl halides (Scheme 1.12). Very often, the addition of a catalytic

    amount of a copper salt dramatically improves the regioselectivity in favor of the 1,4-adduct and

    increases the rate of the reaction.

    MR

    copper catalyst

    MR = RMgBr of RMgCI, ZnR2, RLi

    R = alkyl, aryl

    Scheme 1.12: Generalized copper-catalyzed conjugate addition of organometallic

    carbanions

    The relevant literature on the mechanism of the conjugate addition of organocuprates to

    enones has been reviewed by Woodward38 and Nakamura and Mori39 in 2000. Recently, Minnaard

    and Feringa also reported an article on the mechanism of the copper-catalyzed conjugate addition of

    Grignard reagents to enones.40

    In general, the first step involves the transmetallation of the MR fragment to form an

    organocuprate intermediate that contains the substructure [R2Cu]~. This cuprate fragment binds to

    the "soft" C-C double bond as in 19, whereas the resulting free metal cation M+ (Li+, MgX+, RZn+,

    etc.), being a hard Lewis acid, coordinates to the "hard" oxygen atom. This intermediate has been

    directly observed by low-temperature NMR spectroscopy.41 In the following step, one alkyl group

    is transferred to the ß-position of the conjugated system (Scheme 1.13).

    MNR-Cu-R

    ,MO O'

    , SRpu--j| _CuR

    R

    19

    Scheme 1.13: General mechanism of 1,4-cuprate addition to enones

  • 14 1 Introduction

    The actual mechanism of alkyl group transfer from the copper complex to the enone is still

    matter of discussion. However, two mechanistic possibilities have been proposed. One mechanism

    involves the formation of the Cu(III)-peralkyl intermediate 20 by formal oxidative addition,

    42-44followed by reductive elimination to the adduct 22.

    "

    Alternatively, a 1,2-migratory insertion of

    the alkyl group that leads to the carbocuprated complex 21, followed by a rearrangement to the n-

    adduct 22, is also a viable hypothesis (Scheme 1.14).45

    Ai

    R.\

    /,Cui—

    19

    -CuR

    22

    R 21

    Scheme 1.14: Possibilities of the C-C bond forming step in the cuprate conjugate addition

    The importance of the activation of the enone by the Lewis acid is fundamental. This is

    proved by the fact that the cuprate salt [Li(12-crown-4)2][CuPli2] 23 containing a coordinatively

    saturated Li+ does not react with enones, whereas the corresponding cuprate reagent 24 with free

    Li+ as counterion reacts readily.38

    ^~9'A

  • 1 Introduction 15

    covalent auxiliaries,26,46 the attention of scientists turned to the development of a catalytic version

    using chiral ligands in substoichiometric amounts.7"

    Phosphorus ligands are by far the most widely investigated chiral source. Any source of

    trivalent phosphorus (phosphines, phosphites, phosphorami dites, phosphonites) has been found to

    strongly accelerate the reaction.50 The ideal ratio has been found to be 2 equivalents of P per Cu.47

    A very large amount of chiral bidentate phosphine ligands has been tested in the

    Cu-catalyzed conjugate addition of dialkylzink to enones along with various P,N- and tetradentate

    ligands bearing a P,N,N or P,N,S donor-atom set.47 Moreover, great contributions to this chemistry

    have been achieved with the application of chiral monodentate phosphonite, phosphinite, and

    phosphoramidite ligands derived from biphenol, binaphthol, or TADDOL (Figure 1.7).

    P-R

    0

    ( P-OR'(NR'2)V0

    ( P-OR'(NR'2) ( P-OR'(NR'2)

    Figure 1.7

    An example of copper-catalyzed 1,4-additions of diethylzinc to 2-cyclohexenone with a

    chiral phosphoramidite ligand is illustrated in Scheme 1.16.51

    Et2Zn

    toluene, -30°C

    L =

    Cu(OTf)2 (0.5 mol-%) \/\/ f^TÏT0 )~*L (1 mol-%)

    I I II

    >98 % ee

    Scheme 1.16: Example of highly enantioselective copper-catalyzed conjugated addition

    1.2 Michael Addition

    1.2.1 Discovery and Definition

    The first example of conjugate addition of diethyl sodium malonate to diethyl

    ethylidenemalonate was reported by Komnenos in 1883. However, this chemistry really began a

    few years later in 1887 with the work of Arthur Michael. In his first publication, he reported his

    study on the addition of sodium malonate diethyl ester to cinnamic acid ethyl ester.53 In his

    observations he proposed that this reaction could be applied as a general synthetic method for the

    formation of new carbon-carbon bonds.

  • 16 1 Introduction

    O 9 £ 1)EtOH, reflux, 6h J Jph^^A0B + EtCf^f^OEt 2)h2o "etO^Sr^OEt

    H PhAMichael Michael ^LappBn(m H„r

    O OEtacceptor donor

    Scheme 1.17: First conjugate addition reported by A. Michael

    Since then, this reaction attracted very much attention from the scientific community and

    many efforts have been devoted to its study and development. It has been found that the general

    concept of nucleophilic addition to enones could be applied to a large variety of compound classes

    and is generally accepted as one of the most useful constructive methods in organic synthesis.54

    For such a broad range of reactions, it is appropriate to define a convention of terms.

    "Conjugate addition or 1,4-addition" refers to the addition of any class of nucleophiles, including

    reactive carbanions, to an unsaturated system in conjugation with any electron withdrawing group

    such as those of Chapter 1.1.3. "Michael addition or Michael reaction" refers to the addition of

    stabilized carbanions to unsaturated systems in conjugation with carhonyl groups.

    In organic synthesis, the Michael addition is performed usually by deprotonating the acidic

    ot-carbon with a strong base such as an alkali metal, alkoxide, or amine. After the addition of the

    enone, the product is obtained by hydrolysis with water. To favor the proton exchange, the reaction

    often requires the use of protic solvents such as EtOH. The generalized reaction mechanism is

    depicted in Scheme 1.18.

    Base 9" 9 Q 9~ H+Ri __ >^r1 + ^^ —- A^^K —"R2

    H r2 R1 R2 R1 R2

    31 32

    Scheme 1.18: The generalized mechanism of the base-catalyzed Michael addition

    Unlike other nucleophilic C-C bond forming reactions such as the aldol condensation, which

    require a stoichiometric formation of the enolate, in the Michael reaction the base can also be used

    in catalytic amounts. This allows for milder reaction conditions and, thus, better tolerance toward

    sensitive groups. When a catalytic amount of base is used, every reaction step becomes reversible

    and the reaction proceeds with thermodynamic control of the enolate formation. This is because the

    intermediate enolate species 32, generated by the nucleophilic attack to the ß-carbon, is more basic

  • 1 Introduction 17

    than the original enolate compound 31. The protonation of 32 under the reaction conditions drives

    the reaction to completion.

    1.2.2 Asymmetric Michael Addition

    The most appealing feature of the Michael addition is that, with the appropriate substituents,

    it leads to the generation of one or two stereogenic centers. As the formation of quaternary carbon

    centers is a very attractive achievement from a synthetic point of view, considerable efforts have

    been committed to the development of efficient diastereo- and enantioselective Michael reactions.

    There are essentially three different strategies to control the absolute configuration of the new

    stereocenter in a general conjugate addition (Scheme 1.19).

    diastereo selective

    reactions

    Chiral catalyst R^EWG^

    » R^EWG

    Scheme 1.19: Strategies for the stereoselective conjugate

    The first two approaches, which represent the earliest attempts of stereocontrol in this

    reaction, involve the use of chiral reaction partners. The corresponding product contains, thus, at

    least two stereogenic centers and the reaction proceeds diastereoselectively. The use of chiral

    auxiliaries, which are covalently bonded to the substrates and cleaved at the end of the reaction,

    belongs to this group.

    On the other hand, when the asymmetric element is located on the catalyst, the optically pure

    component, which is usually the most valuable, can be employed in a substoichiometric amount.

    Among the enantioselective Michael additions, it is possible to distinguish further classes

    with respect to the site at which the stereogenic center is generated. In Type I, the stereogenic center

    is formed on the oc-carbon of the nucleophile, whereas in Type II it is formed on the Michael

    acceptor side (Scheme 1.20).55

    Nu

    Chiral conjugate acceptor p'^^^W^ , ^

  • 18 1 Introduction

    Type l

    Z_,H

    R

    C02Et +

    ^Z'

    0

    R H

    ^^C02Et

    Z^H + R^^Z' Z^M_Z'Z H

    Type II

    Et02C^/C02Et +PK

    Ph O

    EtQ2C

    C02Et

    Scheme 1.20: Types of enantioselective Michael addition

    1.2.3 Asymmetric Catalytic Michael Addition

    The function of the catalyst is also an important factor to classify Michael reactions. For

    Type A reactions, the catalyst activates the Michael donor and generates a chiral nucleophile. In

    Type B reactions instead, the catalyst interacts with the Michael acceptor and directs the side of

    approach of the nucleophile. Recently, some catalysts have been called bifunctional to underline the

    simultaneous activation of both nucleophile and electrophile. As we shall see, this is not such a rare

    condition as it is claimed to be.

    Type A

    RO OR

    Type B

    R02C

    R02C

    Scheme 1.21: Types of activation in the catalytic Michael addition

    Catalytic reactions of Type A correspond typically to a stereocontrol of Type I. To this group

    belong all the basic catalysts (amine, phase transfer catalysts, alkoxide, ect...), most transition and

    rare-earth-metal-catalysts, and the proline-derived catalysts that follow the enamine pathway (see

  • 1 Introduction 19

    below). On the other hand, reactions of Type B correspond to a stereoselectivity of Type II and are

    catalyzed by certain transition-metal complexes and proline-derived compounds that follow the

    iminium pathway.

    Basic Catalysts

    After the first report by Bergson,36 who investigated the Michael addition catalyzed by the

    chiral amine 33, Wynberg57 studied a series of cinchona alkaloids as chiral catalysts and found that

    (-)-quinine (34) promoted the Michael addition of ethyl indanone-2-carboxylate to methyl vinyl

    ketone affording the product in excellent yield and with 76 % ee (Scheme 1.22).

    34(1 mol-%)

    CCI4, -21"C

    6h

    99 % yield, 76 % ee

    OH

    33 34

    (-)-quinine

    Scheme 1.22: Michael addition catalyzed by the chiral base

    The Michael addition catalyzed by chiral amines has been developed since. In particular,

    cinchona alkaloids derivatives, in which the nitrogen atom is alkylated to give a chiral salt, have

    been used as phase transfer catalysts.58

    Achiral Bases with Chiral Crown Ethers

    An interesting variation of the base-catalyzed Michael addition was reported by Cram in

    1981. Instead of using chiral bases, he showed that asymmetry is induced by achiral potassium

    bases such as KO'Bu or KM-I2 in combination with chiral crown ethers such as 35, which acts as

    chiral host (Scheme 1.23) for the potassium cation.59

  • 20 1 Introduction

    ^=^ 36 35

    Scheme 1.23: Asymmetric Michael addition with chiral crown ether as K+ host

    The crown ether is coordinated to the potassium-enolate intermediate 36 after the

    deprotonation of the acidic a-proton and provides the necessary asymmetric environment to afford

    excellent enantioselectivity for the substrate tested, 5h (up to 99 % ee). This strategy has been

    further developed, including the use of elaborate sugar-annulated crown ether derivatives.60"63

    However, these systems are characterized by a narrow reaction scope, being effective only with

    specific substrates.

    Proline Catalysts

    After numerous accounts on the use of proline as chiral ligand in asymmetric transition-

    metal-catalyzed reactions, it has been discovered that proline itself can be a powerful

    organocatalyst. Many aspects of the proline-catalyzed reactions have attracted the interest of the

    scientific community. Apart from the obvious reasons that proline is an abundant chiral molecule, is

    very inexpensive, and is available in both enantiomeric forms, the most appealing feature in the

    proline structure is the simultaneous presence of a base and of a Bransted acid. A review of various

    asymmetric reactions catalyzed by proline was published by List in 2002.64

    The modes of reaction are very similar for all the proline-related catalysts and can be divided

    in two groups according to the nature of the reactive intermediate, which is either an iminium ion or

    an enamine, as illustrated in Scheme 1.24.

  • 1 Introduction 21

    iminium pathway N co2H enamine pathway R^/^%.TypeB Type A V^^C\EWG

    R' V_ Nu"^

    Scheme 1.24: Proline catalysis, the two reaction pathways

    In 1991, Yamaguchi65 began his study on asymmetric Michael additions of malonates to a

    series of a,ß-unsaturated ketones catalyzed by L-prolinate salts of alkali metals. The best reactivity

    and enantioselectivity were observed with the rubidium salt 37 as illustrated in Scheme 1.25.66'67

    Unfortunately, prolinate salts were only moderately active and, except the example in Scheme 1.25,

    the extent of asymmetric induction was quite low.

    POzR,

    COzR-i

    Os.

    37 or 38 (5-10 mol- %) R2 O

    R3 Ri02Cy^Ar/

    /—N

    /"n C02HPh H

    COzRt

    37: up to 62 % yield, 77 % ee

    38: up to 93 % yield, >99 % eeN C02RbH

    37 38

    Scheme 1.25: Enantioselective Michael addition catalyzed by proline-related

    organocatalysts

    Noteworthy improvements were obtained by J0rgensen,68 who used the imidazolidine 38 as

    organocatalyst. This molecule, prepared from phenylalanine, was of more general applicability and

    afforded the desired Michael adducts in excellent yield and enantioselectivity.

    In the proposed mechanism, the catalyst reacts with the carbonyl compound to give the

    iminium intermediate 39 shown in Scheme 1.26. The nucleophilicity at the ß-carbon of the iminium

    ion is increased as compared to the a,ß-unsaturated carbonyl. At the same time, the asymmetric

    environment directs the approach of the nucleophile onto the Sï-face (Scheme 1.26).

    75

    Nu

    SI face approach

    Scheme 1.26: Mode of reaction of imidazolidine catalyst involving an iminium pathway

  • 22 1 Introduction

    Some attempts to use proline for the Type B Michael addition were also made. Kozikowski

    and Momose70 reported the intramolecular Michael additions of unactivated ketones to a,ß-

    unsaturated enones catalyzed by proline. However, the reaction required prolonged reaction times

    and a stoichiometric amount of proline. Despite that, the highest enantioselectivity observed was

    34 % ee. Later, List71 and Enders72 applied this methodology to the intermolecular addition of

    unactivated ketones to nitroolefins and obtained excellent diastereo- and good enantioselectivities

    (Scheme 1.27).

    NO;

    +

    R i Ro R'

    ^N^C02HH

    9 R3

    (>15 mol-%) R1 R2

    up to 74 % yield,

    16:1 dr, 76 % ee

    Scheme 1.27: Asymmetric Michael addition of ketones to nitroolefins catalyzed by proline

    Cobalt/Chiral Diamine

    The first example of transition-metal-catalyzed Michael addition was reported by

    Watanabe "' in 1980. In his paper, he reported the clean formation of the Michael adduct from

    nitromethane and calchone catalyzed by a series of Co(II) and Ni(II) salts coordinated to various

    achiral amines.

    In 1984, Brunner75 extended this methodology by using chiral diamine ligands coupled with

    [Co(acac)2] to attain the first asymmetric transition-metal-catalyzed Michael addition. The degree of

    asymmetric induction was moderate and markedly dependent on the structure of the 1,3-dicarbonyl

    compound. The best performances were obtained with the chiral diamine 40 and the indanone

    derivative 5h (Scheme 1.28), however, with moderate asymmetric induction (66 % ee). Treatment

    of [Co(acac)2] with a stoichiometric amount of chiral diamine 40 in toluene afforded the octahedral

    complex [Co(acac)2(40)] (41) as observed by means of MS analysis. Following this

    observation, he proposed that the active intermediate is the cobalt complex 41 containing

    two coordinated enolates of the 1,3-dicarbonyl substrate.

  • 1 Introduction 23

    OEt

    5h

    NH2

    NH;

    40

    [Co(acac)2] (3 mol-%)

    40 (3 mol-%)^

    Toi, -50, 64 h

    >P •*H2

  • 24 1 Introduction

    L =^T "n Ph

    OH

    42

    Q o [Cu(L)]2'H20

    (1-10 mol-%)

    CCI4, 2-3 d

    L = 42: rac.43: 50 % ee

    44: 70 % ee45: 70 % ee

    Scheme 1.29: Copper-catalyzed Type I Michael addition reported by Desimoni

    The mechanism suggested is analogous to that of Brunner's system. Upon addition of the ß-

    keto ester to the copper(II) catalyst, the octahedral complex 46 is formed, in which the tetradentate

    ligand has A-configuration. The attack of the enone was again thought to be directed onto the Re

    face of the substrate by assistance of hydrogen-bond interactions with the equatorial alcohol group

    (Figure 1.8).

    46

    Figure 1.8: The reactive intermediate proposed in the Cu-catalyzed Michael addition

    Recently, J0rgensen reported the first Michael additions of cyclic 1,3-dicarbonyl compounds

    to 2-oxo-3-butenoate esters catalyzed by copper(II)/bisoxazoline (BOX) complexes; this is a rare

    example of a metal-catalyzed Type B Michael addition.77 This methodology afforded the Michael

    adducts in high yield and high enantioselectivity (Scheme 1.30).

  • 1 Introduction 25

    [Cu(OTf)2(BOX)] 0H ph o(10mol-%)

    C02Me C&+C02Me

    BOX =

    98 % yield, 84 % ee

    ,0-^\-Q

    lBu tBu pu~Q tBu

    approach

    Scheme 1.30: Copper/BOX-catalyzed Michael addition

    The structure of the cyclic 1,3-dicarbonyl unit does not allow the bonding in a bidentate

    fashion to the metal center as in the acyclic analogues. Therefore, it has been proposed that the

    copper binds to the enone to form the intermediate compound depicted in Scheme 1.30. This model

    also accounts for the observed absolute configuration of the Michael products: in the structure of

    the proposed intermediate, the S7-face of the double bond is shielded by one tert-buty\ group of the

    ligand whereas the i?e-face is open for the nucleophile approach.

    However, cyclic 1,3-dicarbonyl compounds are completely enolized and their further

    activation is not needed. There is no report of a similar reaction with less enolized 1,3-dicarbonyl

    compounds.

    Shibasuki 's Catalysts

    Shibasaki contributed to the development of the Type II enantioselective Michael addition

    (Scheme 1.20, page 18) with a new kind of catalysts based on heterobimetallic-BINOL systems.

    The first generation of such complexes was reported in 1995 and was based on rare earth metal-

    alkali metal-BINOL complexes. The most active of these was complex 47.78 The catalysts of the

    second generation, with the general formula M[A1(B1N0L)2J where M is an alkali metal (M = Li:

    48), were published one year later and contained aluminum as metal center.79

    The most interesting feature of these catalysts is the simultaneous presence of a basic site—

    the oxygen atom of the BINOL—and two Lewis-acidic sites—the metals. The performances of

    these complexes represent the state of the art for this kind of reaction (Scheme 1.31).

  • 26 1 Introduction

    Bn02C^^C02BnT +

    R

    Na

    '{ ^.Lav ^Na^0 I O

    Na *

    47 or 48(10mol-%)^

    0°C, 24 h

    OaO

    47 and R = Me:

    up to 91 % yield, 92 % ee

    .„^C02Bn 48 and R = H:lc02Bn UP to 88 % y^'d' " % eeR

    (OH

    OH(R)-BINOL

    47 48

    Scheme 1.31: Shibasaki's bifunctional catalyst for Type II asymmetric Michael additions

    The aluminum-catalyzed reaction reported in Scheme 1.31 is so efficient that it has been

    possible to scale up the synthesis to 1.3 kg product still with outstanding chemo- and

    stereoselectivity (91 % yield and 99 % ee).80 NMR spectroscopic investigations and molecular

    mechanics calculations (UFF) disclosed the mode of reaction and the origin of the asymmetric

    induction. On their basis, a similar reaction mechanism has been proposed for both complexes,

    which involves the reactive intermediates 49 and 50, respectively (Scheme 1.32).

    Lewis acidic sites

    49

    50

    Na

    "Na

    R02C

    V

    C02R

    Scheme 1.32: Reaction intermediates showing the basic and Lewis-acidic sites

    After deprotonation of the 1,3-dicarbonyl compound by one oxygen of the BINOL ligand,

    the alkali metal ion coordinates the newly-formed enolate, while the enone binds to the Lewis-

  • 1 Introduction 27

    acidic metal center. The close proximity at which the substrates are kept facilitates the Michael

    addition and this can probably explain the high selectivity even when the reaction is performed at

    room temperature.

    Interestingly, with the second generation catalyst, it was possible to trap the aluminum

    enolate intermediate by addition of an electrophile such as an aldehyde. Thus, stirring enone 51,

    malonate derivative 52, and aldehyde 53 in the presence of the aluminum complex 48, it was

    possible to obtain the product of the tandem Michael-aldol reaction (54) with excellent

    stereoselectivity (Scheme 1.33).79

    EtO OEt Ph' H

    50(10mol-%)

    rt, 36 h

    51 52

    I ?H

    \^C02Et

    53 Ac02Et

    single diasteroisomer

    64% yield, 91 % ee

    Scheme 1.33: Asymmetric tandem Michael-aldol reaction catalyzed by aluminum complex

    48

    In 2000,81 Shibasaki reported a stable and storable modification of his catalytic system (55)

    that features a tetradentate ligand obtained by linking two B1NOL units. Interestingly, the

    protonated modification revealed the highest catalytic activity and selectivity ever obtained with

    such systems (Scheme 1.34).

    + R

    P°2R 55 (10mol-%;

    C02R

    n = 0-3 R^H.Me

    R2 = Me, Bn

    ,C02R2

    Ri C02R2

    82-98 % yield

    £ 98 % ee 55

    Scheme 1.34: Third generation Shibasaki's catalyst for Michael additions

    The methodology developed by Shibasaki is ineffective in the Type I Michael reactions,

    indicating that the catalyst efficiently orientates the enone but does not provide any enantioface

    discrimination for the 1,3-dicarbonyl compound. However, the bifunctional character of the catalyst

    proved to be a successful strategy.

  • 28 1 Introduction

    Ikariya 's Ru/Bisamido Catalyst

    A mechanistically interesting example of Type II Michael reaction was reported by Ikariya

    in 2003.82 In fact, the reaction of dimethyl malonate with 56 formed a rare example of C-bound

    enolato complex in which the 1,3-dicarbonyl compound is deprotonated by the basic amido group.

    The structure of the unusual compound 57 was proven by X-ray crystallography. However, the

    C-bound enolato complex is in equilibrium at room temperature with the free substrate (Scheme

    1.35). Although there was no evidence of an oxygen-bound enolate, the authors did not exclude its

    existence.83

    Tsi

    C02Me Phs^N^ ^K+ \ 1 Ru„

    C02Me Ph // ''y-COjMe

    O

    56 57

    Scheme 1.35: Equilibrium proposed by Ikariya in the formation of the ruthenium complex

    57

    The Michael additions catalyzed by this complex afforded the products with excellent

    enantioselectivity (Scheme 1.36).

    Ms i

    -V* OA

    C02Me+ \

    C02Me

    53 (1 mol-%).0

    tBuOH,60°C,24h y-C02MeMe02C

    99 % yield, 97 % ee

    Scheme 1.36: Ikariya's catalytic Michael addition

    This reaction is special also because it is the only example of Michael addition that is Type

    II with respect of the position of the generated stereocenter, and Type A because the catalyst

    activates the 1,3-dicarbonyl compound.

    Complex 56 was originally employed by Ikariya and Noyori in the asymmetric transfer-

    hydrogenation of ketones.84"92 In analogy with the author's observation in asymmetric

  • 1 Introduction 29

    hydrogénation, they proposed that the N-H moiety actively participates to the reaction by fixating

    the enone in the optimal position by means of a hydrogen bond (Scheme 1.37). A similar hydrogen

    bond participation from a coordinated amine has been previously described by Brunner (see

    Scheme 1.28, page 23).75

    C02Me

    \

    C02Me Ph^N^

    Ph N

    H

    Ph" // '"'r-C02MeH H- //—OUe

    0

    iyis

    Ptf1 N ""

  • 30 1 Introduction

    product of the Michael addition. Thus, this method allowed the direct preparation of chiral alcohols

    with high diastereo- and enantioselectivity.

    Sodeoka's Palladium/Hydroxo/BINAP Catalyst

    Surprisingly, after the preliminary report by Brunner, the use of metal-based Lewis acid

    catalysts for the Type A-I asymmetric Michael addition (that is, with both activation and

    stereocontrol on the Michael donor) has not been further developed until recently.

    Consiglio discovered complexes 59a and 60a and their interconversion depending on the

    acidity of the medium (Scheme 1.39)94

    Ph2 H Ph,~I2

    Pd Pd

    POP

    Ph2 H Ph2

    59 a

    +HBF4

    +H20

    -HBF4

    -H20

    Ph2P\ /OH2

    P OH2Ph2

    60a

    ~I(BF4);

    Scheme 1.39: Acid-base relation between complexes 59a and 60a

    Sodeoka used chiral modifications of such complexes bearing a BINAP ligand (59b-c and

    60b-c, Scheme 1.40) in asymmetric catalysis. After the development of enantioselective

    95-97

    Mukaiyama aldol and Mannich reactions,"

    these complexes were applied to the asymmetric

    Michael addition of diketones and ß-keto esters to a,ß-unsaturated ketones.

    5i

    59b

    \

    H

    Pd Pd

    POP

    H

    59

    ~l (OTf)2

    61b

    ~l (OTf)2

    P\ /OH2

    /dsP OH2

    60

    6i

    60b

    b;Ar=4-Me-C6H4c: Ar= Ph

    Scheme 1.40: Pd/BINAP-enolate formation from hydroxo or water complexes 59 or 60

  • 1 Introduction 31

    It has been observed that the reaction of either 59b or 59c with 1,3-diketone 5i formed

    smoothly the corresponding enolate adducts 61b-c, as indicated by NMR spectroscopy and ESI MS

    analysis. Conversely, the diaqua complex 60b reacts only partially with diketone 5i and the

    formation of the enolato complex 61b is not quantitative.

    Complexes 60b and 60c were tested in the Michael addition of ß-keto esters to methyl vinyl

    ketone. Both complexes showed a similar reactivity with substrate 5i in CH2CI2 at -20°C, but 60b

    gave a better enantioselectivity. Further studies revealed that the reactions carried out in THF

    afforded the products with higher enantioselectivity than those performed in other solvents such as

    CH2C12. With optimized reaction conditions, the Michael adduct 6a was isolated in 92 % yield and

    92 %ee (Scheme 1.41).

    9 O \60b (5 mol-%)92% yield92 % ee

    5a

    THF, -20°C, 24 h \—/ ''—

    6a

    Scheme 1.41

    Interestingly, the bis(hydroxo)-bridged complexes of type 59 were found to be much less

    reactive than the bisaqua-species 60 giving only 28 % of the Michael adduct after 1 week at -10°C

    and 100 hours at 0°C in 66 % ee.

    The screening of other substrates confirmed the broad scope of this reaction, and many ß-

    keto esters with different structural features were converted to their corresponding Michael products

    with more than 90 % ee. Moreover, the reaction showed good enantioselectivity with rarely

    investigated substrates such as a,ß-unsaturated aldehydes and 1,3-diketones (Scheme 1.42).

    60b (5 mol-%)

    H MeOH,0°COMe

    (4 equiv)MeO

    31-73 % yield, 80-87 % ee

    R2

    +60b (10 mol-%)

    !

    THF, -10°C

    (2 equiv)

    up to 90 % yield, 90 % ee

    Scheme 1.42: Michael reactions with unusual substrates

  • 32 1 Introduction

    Following the good performances of the previous reactions, ß-substituted enones were used

    to test the palladium catalysts in the diaslereoseleclive Michael addition. The Michael products

    were formed in high yield (80-98 %) and enantioselectivity (> 93 % ee). The diastereoselectivity

    was generally low (1.2-3.6:1), except for 3-penten-2-one (62), with which a diastereomeric ratio of

    8:1 was observed (Scheme 1.43).

    60b (5 mol-%)

    THF, -20°C, 24 h

    5a 62

    89% yield, dr = 8/1,99%ee

    Scheme 1.43: Diastereo- and enantioselective Michael addition with the Pd aqua complex

    60b

    Interesting insights into the reaction mechanism have been gained by investigating the

    reactivity of 59 and 60 in stoichiometric reactions. It has been observed that the enolato complex

    61, formed in situ from the reaction of 59 and 5i, does not react with methyl vinyl ketone (2 equiv)

    in THF solution. However, addition of triflic acid (1 equiv) to this solution triggers the reaction to

    give the Michael adduct in 96 % yield and 97 % ee. The same experiment was carried out starting

    from 60; after addition of the 1,3-diketone 5i, the NMR spectrum showed an equilibrium between

    60, 5i, and 61. The addition of methyl vinyl ketone to this mixture afforded smoothly the Michael

    product without the need of TfOH in comparable yield and enantioselectivity (99 %, 91 % ee).

    H

    f\ A SPd Pd

    POPH

    59

    ~l (OTf)2

    5i

    "lOTf Ol(2 equiv)

    X-

    + TfOH (1 equiv)

    no reaction

    6i

    61

    tR„~IOTf

    P\ /OH2Pd

    P OH2

    60

    "I (OTf)25i

    Ol(2 equiv)

    61 6i

    Scheme 1.44: Stoichiometric reactions with Pd-enolato complexes

  • 1 Introduction 33

    This reactivity was observed with ß-keto ester 5a as well, except that the amount of enolato

    complex was significantly reduced in the case of the reaction of 5a with 60b.

    Consequently, Sodeoka proposed that the palladium enolato complex was an important

    intermediate in the reaction mechanism. Since the addition of TfOH started the reaction but also

    triggered the formation of the equilibrium between 60 + 5i and 61, it was unclear whether the

    diaqua complex 60 also acts as Lewis acid co-catalyst and activates the enone. However, this

    possibility was ruled out by kinetic experiments, which showed that the reaction was first-order

    with respect to the catalyst. Therefore, it has been proposed that TfOH itself, generated upon

    coordination of the 1,3-dicarbonyl substrate, activates the enone by protonation. The proposed

    reaction mechanism (Figure 1.9) will be discussed in detail as it is relevant to the results exposed in

    this thesis.

    M*.

    Figure 1.9: Catalytic cycle of Sodeoka's Michael addition

    Although complex 59 was not observed under the catalytic conditions, Sodeoka assumed its

    formation from 60 based on the reactivity discovered by Consiglio. Complex 59 binds the ß-keto

    ester in a bidentate fashion while the hydroxo group abstracts the acidic cc-proton from the

    coordinated 1,3-dicarbonyl moiety to form the enolate 6161. However, the authors did not rule out

  • 34 1 Introduction

    the possibility that the Lewis-acidic activation of the 1,3-dicarbonyl compound by coordination to

    palladium allowed the acidic proton to be abstracted by a weak base such as a water molecule. At

    this point, the triflic acid would protonate the enone which would promote the nucleophilic attack of

    the coordinated enolate fragment. After tautomerization of the unstable enol intermediate and ligand

    substitution, the cycle was terminated.

    1.3 Heterofunctionalization of 1,3-Dicarbonyl Compounds

    In recent years, the "soft" enolization of 1,3-dicarbonyl compounds by complexation with

    Lewis acids as discussed in Paragraph 1.1.3 has permitted the development of catalytic processes

    other than the Michael addition. This paragraph describes some examples of asymmetric

    heterofunctionalizations of 1,3-dicarbonyl compounds catalyzed by chiral Lewis acids.

    In 2000, the first catalytic asymmetric fluorination of ß-keto esters was discovered in our

    laboratory.99"103 By using the titanium/TADDOL complex 25 as catalyst, various 1,3-dicarbonyl

    compounds have been successfully converted to their fiuorinated analogues with excellent yields

    and enantioselectivities up to 90 % ee (Scheme 1.45). Selectfluor® (also called F-TEDA) has

    proved to be the best source of electrophilic fluorine.

    a. °-

    25NP>< Cl\NPNp CK^OMeCN^I SNCMe

    CI

    F ~l(BF4-)2

    O O ^Nt O O

    ° V MeCN.rt *OV°2

    CH2CI

    F-TEDA up to 90 % ee

    Scheme 1.45: Ti-catalyzed enantioselective electrophilic fluorination of ß-keto esters

    In a first mechanistic rationalization, it has been proposed that the reaction involves the

    coordination of the ß-keto ester to the titanium complex in a bidentate fashion upon displacement of

    one chloride and one MeCN molecule, followed by deprotonation to form an enolato

    intermediate.104 The Ti-enolato complex is attacked by the electrophilic F-TEDA with high

    stereoselectivity. The proposed catalytic cycle is illustrated in Scheme 1.46.

  • 1 Introduction 35

    [TiCI2(TADDOL)(MeCN)2]

    25

    5d-HCl

    -MeCN

    Ö

    26

    O

    1"NPy>^>Np1-Np^\ CI 7^1-Np

    MeCN | O ,

    5d1-NPv °V ? 1-Np

    1-Np

    F-TEDA

    MeCN

    Scheme 1.46: The proposed mechanism of the titanium-catalyzed asymmetric fluorination

    Starting from these hypotheses, QM/MM investigations have been carried out to better

    understand the reaction mechanism.104 The calculations showed that the C-F bond formation step

    involves a single electron transfer from the complex to F-TEDA. The origin of the asymmetric

    induction was investigated by minimizing the energy of the structures of all the possible Ti-enolate

    intermediates. The most stable diasteroisomer displayed one face-on naphthyl group of TADDOL

    shielding the Re enantioface of the ß-keto ester. Therefore, F-TEDA can only approach from the Si

    side. The experimental determination of the absolute configuration of the fluorinated product

    supported this model.

    Recently, attempts to prepare the intermediate enolato complexes to the type

    [TiCl(TADDOL)(enolato)(MeCN)] from the equimolar reaction of 25 with Na or Li enolate of

    some ß-keto esters gave the 1:2 adducts instead of the expected 1:1 complexes.105 These 1:2 adducts

    have been characterized in solution and have been identified as a complex mixture of up to 6

    possible diastereoisomers. Nevertheless, stoichiometric and catalytic reactions with these

    complexes afforded the corresponding fluorinated products with the same sense of induction and

  • 36 1 Introduction

    comparable enantioselectivity as in the fluorination catalyzed by 25.105 Therefore, the identity of the

    actual intermediate is still a matter of discussion.

    This methodology was extended to chlorination and bromination106 (Scheme 1.47) and also

    to the enantioselective geminal diheterohalogenation of unsubstituted ß-keto esters101 (Scheme

    1.48).

    Q O Ph

    0 Ph+ N-X

    [TiCI2(MeCN)2(TADDOL)] 25

    (5 mol-%)

    O 0 Ph

    MeCN

    y\ O Phx

    X

    X = CI: 88 % ee

    Br: 23 % ee

    Scheme 1.47: Chlorination and bromination of ß-keto esters catalyzed by Lewis-acidic

    complex 25

    ^UL

    1) F-TEDA

    2) NCS

    OBn

    65 % yield

    66 % eeCI F

    2 BF4-

    F-TEDA= FNV^N-CH2CI

    [TiCI2(MeCN)2(TADDOL)] 25

    (5 mol-%)

    NCS =

    1)NCS

    2) F-TEDA

    OBnF CI

    60 % yield

    57 % ee

    N-CI

    Scheme 1.48: One-pot geminal diheterohalogenation of ß-keto esters

    Analogously, 25 was also applied in other a-functionalizations of ß-keto esters such as

    sulfenylation and hydroxylation (Scheme 1.49).

  • 1 Introduction 37

    PhSCI

    25(1.5mol-%)

    5b

    84 % yield, 88 % ee

    25

    25(1.5mol-%) 97 % yield, 94 % ee

    -Nz

    28

    OR

    Scheme 1.49: Asymmetric sulfenylation and hydroxylation of ß-keto esters catalyzed by 25

    The proposed mechanism of a-hydroxylation was similar to the Ni(II)-catalyzed

    hydroxylation of ß-keto esters with dimethyldioxirane previously reported by Adam and Smerz.

    The authors proposed that the key step in their reaction involved the epoxidation of the C-C double

    bond of the metal-bound enolate (Scheme 1.50).

    Scheme 1.50: Proposed mechanism of the metal-catalyzed hydroxylation of ß-keto esters

    with dimethyldioxirane

    The dioxirane compound, though, was not the preferred oxidating agent for the titanium

    catalytic system because traces of water, derived from the preparation of the dioxirane, deactivate

    the catalyst and reduced the conversion to the desired product. This problem was solved using the

    non-hygroscopic oxaziridine 28 as oxidating agent.

    This protocol was efficient, however, only for substrates that do not enolize spontaneously

    under neutral conditions such as 5a. The catalytic reaction with highly enolizable 1,3-dicarbonyl

    compounds, such as the a-acyl-ô-valerolactam 5g, afforded the products in almost quantitative

    yield but with limited enantioselectivity (Scheme 1.51). Since these substrates are present in the

    enol form to a significant extent even without a Lewis acid, they react with the oxaziridine 28 in the

    absence of the catalyst, thus explaining the origin of the low enantioselectivity.

  • 38 1 Introduction

    Ph02S

    Y//

    28

    NO, 29

    25 (5 mol-%)

    N Ph

    6g: 66 % enol*

    MeCN, rt

    OH

    97 % yield, 94 % ee

    or

    N Ph

    HO

    97% yield, 19 % ee

    30

    * enol content measured by integration of the representative H NMR peaks in CDCI3

    Scheme 1.51: Ti-catalyzed hydroxylation with substrates having low and high enol content

    Analogous a-heterofunctionalizations of 1,3-dicarbonyl compounds catalyzed by transition

    metal complexes have been reported, among others, by Sodeoka"2 (Scheme 1.52) and

    Jorgensen"3,114 (Scheme 1.53). Similarly to the titanium-catalyzed reactions, in both cases the

    function of the Lewis acid is to promote the formation of the enolate and to provide an asymmetric

    environment for the approach of the electrophile.

    NFSI

    cat (2.5 mol-%)

    EtOH

    cat =

    5b 27: 90% yield, 92% ee

    H

    Pd Pd

    POP

    H

    ~l (OTf)2

    NFSI

    Ph Ph1 1

    02S^ ,so2N

    1

    F

    N Q

    or

    Ar = Ph or 3,5-(CH3)2-C6H3 Ar = 3t5-CBu)2~4-W&0)-C6H2

    Scheme 1.52: Enantioselective fluorination with Pd/BINAP-system reported by Sodeoka

  • 1 Introduction 39

    < T T >

    XX +Bn02C

    C02Bn

    Ph Ph

    (10mol-%)

    R2 /^CC^BnBn02C-NH

    up to 98% yield, 98% «

    R2

    Ri, R2 = alkyl

    R3 = OR or alkyl

    Cu(OTf)2 (10mol-%)

  • 40 1 Introduction

    zinc(II)125-130 (77 = 10.88), ruthenium(II)126''31"133 (77 = 5.86), palladium(II)134 (77 = 6.75) and

    platinum(II)134 (77 = 8.0), and gold(III)135 (77 = 8.4) have been prepared and characterized.1

    The simple preparation from the condensation of ethylene diamine and salicylic aldehyde

    contributed to the popularity of this ligand in coordination chemistry. Additionally, the modular

    synthesis allowed the preparation of a conspicuous amount of analogues, transforming the term

    salen into the representative of a whole class of ligands.

    The most eminent application of these ligands in catalysis is the asymmetric epoxidation of

    olefins developed by Jacobsen137 by using chiral complexes of the type [MnCl(salen)] (such as 64)

    as catalysts. The epoxides were obtained in excellent enantioselectivities (Scheme 1.54).

    Ph+ NaOCI

    64 (5 mol-%)

    CH2CI2

    A

    Ph

    92 % ee

    Scheme 1.54: Jacobsen's epoxidation of olefins catalyzed by salen complex 64

    Initially, Jacobsen and Katsuki rationalized the enantioselectivity with a model in which the

    olefin approaches the metal-oxo bond from the side and parallel to the coordination plane (side-on

    approach). Depen