university of groningen catalytic enantioselective conjugate … · 2016-03-08 · damage.4 the...

University of Groningen

Catalytic enantioselective conjugate addition of organometallic reagentsde Vries, Andreas Hendrikus Maria

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:1996

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):de Vries, A. H. M. (1996). Catalytic enantioselective conjugate addition of organometallic reagents.Groningen: s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 07-09-2020

https://www.rug.nl/research/portal/en/publications/catalytic-enantioselective-conjugate-addition-of-organometallic-reagents(9ef5aa13-4592-4c91-895d-9953750e4902).html

https://www.rug.nl/research/portal/en/publications/catalytic-enantioselective-conjugate-addition-of-organometallic-reagents(9ef5aa13-4592-4c91-895d-9953750e4902).html

Catalytic Enantioselective Conjugate Addition of Organometallic Reagents

Omslag: Martine Hoving (grafisch ontwerper)

Druk: PrintPartners Ipskamp B.V., Enschede

RIJKSUNIVERSITEIT GRONINGEN

Catalytic Enantioselective Conjugate Addition of Organometallic Reagents

Proefschrift

ter verkrijging van het doctoraat in deWiskunde en Natuurwetenschappenaan de Rijksuniversiteit Groningen

op gezag van deRector Magnificus, dr. F. van der Woude,

in het openbaar te verdedigenvrijdag 18 oktober 1996

des namiddags te 4.00 uur

door

Andreas Hendrikus Maria de Vries

geboren op 5 september 1967te Swifterbant (Z.IJ.P.)

Promotor: Prof. dr. B.L. Feringa

Dankwoord

Promoveren doe je gelukkig niet in je eentje! Met dit schrijven tracht ik iedereen te bedanken die binnen en buiten het "lab" heeft bijgedragen aan de totstandkomingvan dit proefschrift. Aangezien je dit leest is de kans groot dat jij tot die mensen behoort die mij het leven op één of andere manier prettig heeft gemaakt. Eenaantal van die grote groep mensen wil ik met name bedanken.Allereerst mijn promoter, Prof. dr. Ben L. Feringa, bedankt voor de begeleiding van hetonderzoek. Ben, jouw grote koffer vol met ideeën en wilde voorstellen en jouw nooitaflatende enthousiasme voor de gevonden resultaten zijn een grote stimulans geweestvoor het onderzoek beschreven in dit proefschirft. Verder schept jouw voortdurendedrang om in vele zaken - chemische maar ook vooral niet-chemische disciplines - te willen uitblinken een goede werksfeer en een gevoel van kameraadschap.Onderz oek doen is een continu leerproces waarbij veel wordt opgestoken van ervarenmensen om je heen. Daarom wil ik Onko Jan Gelling, Ben de Lange, Johan Jansen, Wolter Jager en Nico Kiers bedanken voor de vele waardevolle tips en sugg esties tijdens de eerste jaren van mijn (hoofdvak)onderzoek. Verder ben ikJohan Jansen en Ron Hulst zeer erkentelijk voor respectievelijk de eerste schreden ophet gebied van de nikkel gekatalyseerde reakties (hoofstuk 3) en de synthese van hetfosforamidaat beschreven in hoofdstuk 5. De discussie s met collega's met verwante onderzoeksprojecten hebben een belangrijkeinvloed gehad op de voortgang van dit onderzoek; Marcel Lubben, Robert Hof, JoopKnol, Erik Keller, Charon Zondervan, Arjan van Oeveren en Minze Rispens bedanktvoor de interessante en nuttige bespiegelingen aangaande het onderzoek. Esther vanden Beuken en Ron Hulst wil ik bedanken voor de tips over fosforchemie. Verder zijnde bijdragen van Rosalinde Imbos, als hoofdvakstudente verantwoordelijk voor veleexperimenten beschreven in hoofdstuk 4, en Alexander Arnold (Erasmusstudent) onontbeerlijk geweest voor de voltooiing van dit proefschrift.Marc Veen, Rob Zi jlstra en Jan van Esch ben ik dankbaar voor het oplossen van allerleiproblemen aangaande computer- en e-mailzaken. Het hachelijke, maar spannende zeilavontuur met de twee laatstgenoemden is een dierbare herinnering geworden.Mede labbewoners door de jaren heen schiepen een sfeer waar het prettig was om te werken, echter vooral de activiteiten buiten het "lab", zoals binnen- en buitenlandse werkweken en de vele, vaak vreemde, sportieve evenementen zullen mealtijd bijblijven. Alle zaal- en vleugelgenoten creëerden een luidruchtige, maar fijne werkomgeving. Twee speciale zaalgenoten van de laatste jaren wil ik in het bijzonder bedanken. Edzard Geertsema en Jan-Willem Weener, zonder jullie had ik veel minder plezier in het werken gehad.To this extent, I've to thank the two aliens Chiu Leung and Filippo Minutolo as well.

Without our friendship, my English was still like your Dutch.The research visit to the laboratory of Prof. dr. P. Knochel had a great impact on thisthesis. Thanks to all the members of AK Knochel, especially chef Lothar Schwink, forthe excellent guidance to the chemistry of functionalised dialkylzincs and forthe pleasant stay in Marburg.Veel dank ben ik verschuldigd aan: Auke Meetsma, voor de opheldering van een essentiële kristalstruktuur; Marinus Suijkerbuijk, voor de assistentie bij de vele e.e. bepalingen met behulp van chirale HPLC; het NMR-service team, met nameWim Kruizinga en Jan Herrema, voor de e.e. bepalingen waarbij NMR technieken vereist waren; Albert Kiewiet voor de exacte massa bepalingen, en de mensen van de elementenanalyse en het secretariaat, voor bewezen diensten.De leescommissie bestaande uit Prof. dr. A.M. van Leusen, Prof. dr. R.M. Kellogg en Prof. dr. J.H. Teuben ben ik erkentelijk voor de snelle, kritische correcties en de beoordeling van het manuscript.

Vele mensen buiten het "lab" hebben indirect meegeholpen aan de voltooiing van ditproefschrift. Allereerst wil ik mijn ouders, broers, zus en partners bedanken voor dealtijd prettige en levendige thuisbasis en voor het onuitgesproken vertrouwen wat julliein mij hebben. Hopelijk wordt door dit boekje duidelijk wat ik de afgelopen jaren hebgeleerd.De mannen, zijnde Marcel "Loepie" Lubben, Wim "Doccie" Dokter, Marcel "Firestarter" voorheen "Otje" Ottens, Aldert "Vanb" van Buuren, Arnoud "Zwol" Dijkstra en Ivo "Stresser" Staijen, bedankt voor de talloze onvergete-lijke momenten tijdens onze - door anderen verguisde - wetenschap-pelijke bijeenkomsten. Nimmer heb ik de chemie van zoveel kanten gezien. Zonder jullie zou e-mail voor mij net zoveel betekenen als een kleurentelevisie voor de Yanomami-indianen in het Amazone regenwoud.De voetball ers en voetbalsters van The Knickerbockers hebben door de jaren heen eenbelangrijke plaats in mijn leven ingenomen. De unieke sfeer tijdens trainingen en wedstrijden, maar ook vooral in het clubhuis zal ik nergens anders aantreffen. De kleurencombinatie van de omslag moet gezien worden als een ode aan de mooiste voetbalclub van Nederland.Naast de eerder genoemde familie, vrienden en collega's (ook de niet bij naam genoemden) wil ik tenslotte de volgende mensen bedanken voor het geven van plezier, liefde en kleur aan mijn leven: Jan Geert en Martine Dijkstra, Tanja Ku iphuis, Esther Juurlink, Kuna van der Wulp, Suzanne Nieborg, Bertus Aafjes,Jesse de Gooijer, José van Vught, de Risk-slachtoffers: Arjen Bouter en Danny Staal, Peter van der Meer, Frank de Vries, vele andere team- en trainings-genoten, medebestuursleden en leden van The Knickerbockers.

Contents

Chapter 1 Introduction

1.1 Routes to enantiomerically pure compounds . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Catalytic asymmetric carbon-carbon bond formations: selected examples

21.3 Aims of this thesis and survey of its contents . . . . . . . . . . . . . . . . . . . . . . . . .

81.4 References and notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Chapter 2 Carbon-carbon bond formation by catalytic enantioselectiveconjugate addition

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13

2.2 Asymmetric metal-mediated 1,4-addition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.3 Enantioselective conjugate addition of Grignard reagents . . . . . . . . . . . . . . 182.4 Catalytic enantioselective conjugate addition of organolithium reagents . 232.5 Conjugate addit ion of dialkylzinc reagents catalysed by chiral nickel complexes

252.6 Catalytic Michael additions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282.7 Nitroalkane additions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362.8 Miscellaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372.9 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382.10 References and notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

Chapter 3 Conjugate addition of diethylzinc to chalcones catalysed bychiral nickel complexes

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433.2 Preliminary experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443.3 Synthesis of chiral $-amino alcohols derived from (+)-camphor . . . . . . . . . 463.4 (+) -Camphor-derived $-amino alcohols as ligands in the nickel catalyse d

additionof diethylzinc to chalcone. Optimisation of the catalyst composition . . . . . 48

3.5 Variation of substrate, solvent, and reagent . . . . . . . . . . . . . . . . . . . . . . . . . . . 523.6 Mechanistic aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553.7 Summary and concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593.8 Experimental section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593.9 References and notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

Chapter 4 (+)-Camphor-derived tri- and tetradentate $$-amino alcohols;synthesis and application as ligands in the nickel catalysedenantioselective conjugate addition of diethylzinc

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694.2 Synthesis of (+)-camphor-derived tri- and tetradentate

$-amino alcohols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704.3 (+)-Camphor-derived tri- and tetradentate $-amino alcohols

as chiral ligands in the nickel catalysed additionof diethylzinc to chalcone and cyclohexenone . . . . . . . . . . . . . . . . . . . . . . . . 74

4.4 Summary and concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784.5 Experimental section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784.6 References and notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

Chapter 5 Towards a new catalytic system for enantioselective conjugateaddition of organometallic reagents

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85

5.2 The reactivity of diethylzinc towards enones in the presence of metal salts 855.3 Copper catalysed enantioselective conjugate addition of diethylzinc to enones

875.4 Enantioselective copper catalysed methyl transfer to enones with

trimethylaluminium as organometallic reagent . . . . . . . . . . . . . . . . . . . . . . . . 945.5 Summary and concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 975.6 Experimental section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 985.7 References and notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Chapter 6 Novel chiral monodentate phosphorus amidites; synthesis andapplication as ligands in the copper catalysed conjugate additionof dialkylzinc reagents to cyclic and acyclic enones

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056.2 Determination of the molecular structure of the complex of CuI and 6.1 . . 1056.3 Synthesis of novel chiral phosphorus amidites . . . . . . . . . . . . . . . . . . . . . . . . 1076.4 Examination of the modified phosphorus amidites as chiral ligands in the

copper catalysed conjugate addition of diethylzinc to cyclohexenone and chalcone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

6.5 Variation of the reaction conditions, substrate, and dialkylzinc reagent . . 1176.6 Mechanistic aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1206.7 Summary and future prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1236.8 Experimental section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246.9 References and notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Chapter 7 Asymmetric catalysis with phosphorus amidites

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1397.2 Catalytic Michael addition reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1397.3 Copper catalysed S 2' reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140N

7.4 Catalytic asymmetric addition of organometallic reagents to imines . . . . . 1427.5 Addition of diethylzinc to tropone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1437.6 Copper catalysed enantioselective allylic oxidation . . . . . . . . . . . . . . . . . . . 1437.7 Rhodium catalysed enantioselective hydroformylation . . . . . . . . . . . . . . . . . 1447.8 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1457.9 Experimental section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1457.10 References and notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147

Samenvatting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

1

Chapter 1

Introduction

1.1 Routes to enantiomerically pure compounds

The synthesis of optically active compounds is a subject that has fascinated chemistsfor more tha n a century. Since the pioneering work of Pasteur, van 't Hoff and Le Bell, 1

the area of stereochemistry began to evolve into the major field of research that it isnowaday s. In recent years, the synthesis and isolation of enantiomerically pur e2

compounds has gained new impetus, due to the realization that a chiral compoun dinteracts with enantiomers in different ways as a result of a diastereomeri crel ationship. A well known example of the dramatic difference in activity o f3

enantiomers is thalidomide, commercially known as Softenon. Only one enantiomergives the desired therapeutic effect, whereas the other enantiomer causes severe fetaldam age. The search for efficient syntheses of enantiomerically pure compounds i s4

going on, largely stimulated by the requirements for new bioactive materials.In general there are three main routes to obtain pure enantiomers (Scheme 1.1): 5

- resolution of a racemic mixture- synthesis with compounds from the chirality pool- asymmetric synthesisAltho ugh significant advances in other routes to obtain enantiomerically pur ecompo unds have been reported, the 'classical' resolution of racemates b ydiastereomeric crystallisation still constitutes the most important method in industry. 6

Howev er, the maximum theoretical yield of one enantiomer is 50%, unless th eunwanted enantiomer can be recycled. In a kinetic resolution the two enantiomers ofa racemic mixture react at different rates with a chiral entity, preferably used i ncatalytic amounts. Excellent kinetic resolutions have been reported by employin g7

enzymes. 8

Naturally o ccurring chiral compounds (referred to as the chirality pool) can be used asstart ing materials for enantiomerically pure compounds or may be employed a s9

enantios elective agents (catalysts or ligands) in organic synthesis. The lack o f10

availability of both enantiomers of most natural compounds often is a limiting factor.Therefore many desired enantiomers have to obtained by synthesis.In the e arly days, synthesis to enantiomerically pure compounds from prochira lprecursors was considered possible only by using biochemical methods. Althoug hpowerful, t hose methods using enzymes, cell cultures, or (living) microorganisms are11

in most cases substrate specific. Organic synthesis, on the other hand, has revealed12

Racemates

Synthesis

Chirality Pool

Enzymatic Chemical

Diastereomercrystallisation

Kineticresolution

Enantiomerically pure compound

Biocatalysis

Asymmetricsynthesis

ProchiralSubstrates

Catalysis

Chapter 1

2

a variet y of versatile stereoselective reactions that complement biological processes. 13

Optically activ e compounds can be obtained using either a stoichiometric or a catalyticamount of chir al auxiliary. All stereoselective syntheses are based on the principle thatthe products are formed via diastereomeric transition states that differ in Gibbs freeenergy of activation. If this energy difference is sufficient ( $ 3 kcal/mol, RT) on eenantiomer will be formed, preferentially.

Scheme 1.1 Routes to enantiomerically pure compounds.

Asymm etric catalysis is the most promising and attractive form of stereoselectiv esynthesis , since a small amount of enantiomerically pure material produces larg equantities of enantiomerically enriched, or in the ideal situation, enantiomerically puremateria l. A wide variety of highly successful reactions with enantiomeric excesse s(e.e.'s) > 95% have been reported. In most cases chiral transition metal complexes,14

oft en prepared in situ, are employed as the catalysts. The reactions involved ar e15

generally asymmetric reduction, asymmetric oxidation and asymmetric carbon-carbonbond formation. Especially the last category has a tremendous synthetic utility and inthe next Section selected examples of successful catalytic asymmetric carbon-carbonbond for mations will be highlighted. For other supplementary information concerningchirality and (catalytic) asymmetric synthesis, the interested reader is referred to therefere nces throughout this Chapter and to five of the six last Ph.D. theses from th eOrganic Department of this University. 16

(S)-1.1Cl3C

Cl3C COOEt+ N2CHCOOEt

NCu

OO

RR

2 (S)-1.1

R =

C8H17O

t-Bu cis / trans = 85/15e.e. (cis) 91%

Introduction

3

1.2 Catalytic asymmetric carbon-carbon bond formations:selected examples

Asymmetric cyclopropanationAn early e xample of metal-based homogeneous asymmetric catalysis dates from 1966.It involve s an asymmetric cyclopropanation catalysed by a complex of copper (II) anda chiral Schiff base derived from salicylaldehyde and "-phenylethylamine. Although17

enantiom eric excesses were low, the principle was elaborated by Aratani and co -workers and has led to high enantioselectivities in selected cyclopropanatio nreactions. An example is given in Scheme 1.2.18

Scheme 1.2

Recent ad vances in the development of catalysts for enantioselective intra- an dintermolecular cyclopropanation reactions were achieved with dinuclear rhodium (II)complexes of chiral pyrrolidone or oxazolidinone ligands and copper complexe s19

derived f rom a chiral C -symmetric semicorrin ligand 1.2 or structurally relate d220

bis(oxazo line) ligands, for example 1.3 . Some results obtained with these coppe r21

catalysts in the cyclopropanation of styrene with ethyl diazoacetate are compiled inScheme 1.3. The mechanism of these highly selective reactions presumably involvesa cycloaddition of a metal-carbenoid species to the olefin. 18,19

1.31.2

+Ph COOEt

N2CHCOOEt+PhPh COOEt

CuII

1.2 or 1.3

N N

O O

t-Bu t-Bu

NH

N

CN

OH HO99%73 / 27

85%75 / 25

1.3

1.2

trans / cis e.e. (trans)

Chapter 1

4

Scheme 1.3

Asymmetric Diels-Alder reactionsThe Die ls-Alder reaction is a classical reaction in organic synthesis and especially thevery recent progress in the design of chiral Lewis acids has led to the development ofeffective catalysts for asymmetric Diels-Alder reactions. Most efforts were focussed16e

on the use of boron, aluminium, and titanium as the Lewis acidic center, however ,successful cat alysis was alo achieved with catalysts based on copper and lanthanides. 22

Since Knol has recently reported extensively on this subject, only two examples are16e

shown in Scheme 1.4.

1.5

Ph NHSO2CF3

NHSO2CF3Ph

e.e. 96%

e.e. 99%

1.4

OO

O

OB -H+

1.4CHO

+O

H

H

BnO

NO

O

OOBnON

O O+

1.5AlMe3

1) chiral cat.2) H3O

+/ H2O Ph

OHEt2Zn+

Ph H

O(1.1)

Introduction

5

Scheme 1.4 The Brøndsted acid-assisted chiral Lewis acid 1.4 as chiral catalyst fora D-A reaction of ",$-enals with cyclopentadiene and the aluminium /22c

chiral bis-sulfonamide 1.5 catalysed Diels-Alder reaction of a bidentatedienophile and a substituted cyclopentadiene. 22e

Asymmetric 1,2-additionThe addition of carbon-nucleophiles to aldehydes (or other carbonyl compounds) togive seconda ry alcohols is one of the basic reactions of organic synthesis. Asymmetricvariations of this reaction have been developed by addition of dialkylzinc reagents toaldehydes in presence of catalytic amounts of chiral amino alcohols or cinchon a23

alkaloid bases (Eq. 1.1). At present, the 1,2-addition of (functionalised) dialkylzinc15

reagents to aldehydes is one of the most studied catalytic asymmetric transformations,in pa rticular to examine whether a novel chiral ligand is capable of asymmetri cinduction. 24

Anal ogous catalytic enantioselective 1,2-additions to carbonyl compounds wer eachieved with silyl enol ethers and allyl silanes and allyl stannanes, furnishing chiral25 26

e.e. 80 - 98%

1.6

OO

TiCl2

dr. 73 / 27 - 99 / 1e.e. 77 - 99%

R2 COOR

OH

R1

OR'Me2Si

1.6R2

OSiMe2R'

R1

+H COOR

O

Bu3SnR H

O+ 1.6

R

OH

Chapter 1

6

alcohols with p ossibilities to extend the carbon framework. A breakthrough in this areawas achieved with chiral titanium complex 1.6 and examples of highly enantioselectivealdol and allyl transfer reactions catalysed by 1.6 are shown in Scheme 1.5.27 28

Scheme 1.5 Enantioselective aldol and allyl transfer reactions.

Very recently, highly enantioselective allylations of a wide variety of non-enolisablealdehydes (aromatic and unsaturated) with allyltributyltin using catalytic amounts ofBINAP and AgOTf have been reported (e.e. > 88%). 26, 29

Optically active cyanohydrins have been recognized to be versatile compounds sincethey can be easily co nverted to a variety of chiral molecules. Therefore, much effort hasbeen focussed on asymmetric cyanohydrin syntheses using enzymes, chiral cycli c30

dip eptides, chiral Lewis acids, and chiral Schiff bases of dipeptides. Promisin g31 32 33

results were achieved with cyclic dipeptide 1.7 and Schiff base 1.8 in the presence31 33

of Ti(O i-Pr) as catalyst in the addition of hydrogen cyanide to several aromatic and4

aliphatic aldehydes (Scheme 1.6). 34

R = aromatic, aliphatic e.e. 50 - 97%

1.8

N

OH

N COOMeR1

O R2

H

1.7

HNNH

O

O

N

NH

R H

O+ HCN

R CN

OHH

Ti(Oi-Pr)4

1.7 or 1.8

R = Me, n-Bu, vinyl e.e. 79 - 91%

(1.2)N RLi+ 1.3

R

NHH

Introduction

7

Scheme 1.6

Compare d to the addition to carbonyl compounds, successful examples of asymmetricaddit ion of carbon nucleophiles to imines has been reported only scarcely. Tw ocontributions n eed to be emphasised. Nucleophilic additions of organolithium reagentsto N-arylimines ar e promoted by C -symmetric bis(oxazoline) ligands (for example 1.3)2

with high asymmetric induction and decreasing the ligand loading to 20 mol% retainsmost of the ceiling e.e.'s (Eq. 1.2). 35

Another very recent communication illustrates an effective catalytic Strecker synthesis.Employing cyc lic dipeptide 1.9 as catalyst, the conversion of aldehydes to amino acidsprecursors occurs in high yield and in some cases exceptionally high e.e.'s (Scheme1.7). The basic guanidine side chain, accelerating proton transfer in the Strecke r36

synthesis, was es sential for asymmetric induction since the structural resembling cyclicdipeptide 1.7 failed to afford any enantioselectivity.

R = benzyl

1.9

HNNH

O

O

N

NH

NH2H

e.e. 80 - 99%

+ HCN 1.9

NR

X

NR

CN

H

X

Chapter 1

8

Scheme 1.7

Asymmetric allylic alkylationThe ea rliest examples of palladium catalysed enantioselective allylic alkylation screating a stereogenic center at the allylic or homoallylic position, were achieved37 38

employing chiral phosphine ligand 1.10 (DIOP) (e.e. 22-46%). Although enantiomericexcesses were low t o moderate, this principle had lead to the synthesis of a wide varietyof bis phosphines and to successful application in the allylic alkylation to acycli csubstrates (e.e. 30-90%). 39

Recent advances in the catalytic enantioselective allylic alkylation of the model acyclicsubstra te (R' = phenyl) have been achieved using C -symmetric ligands, such as chiral2

sem icorrin 1.2 and bis(oxazoline) 1.3 , with enantioselectivities up to 97% (Schem e1.8). Ligand 1.11 (Ar and Ar = Ph; R = i-Pr or Ph), representing a new class o f40

1 2 1

ligands, showed record enantioselectivities not only for the substitution of the modelacyclic subtrate (e.e . up to 99%), but for the first time also with 1,3-dialkyl-2-propenyl41

acetates (R' = alkyl). 41a

1.11

O

NR1

P Ar1

Ar2

e.e. 22 - 99%

1.10

O

O PPh2

PPh2

H

H

R' R'

C(CO2R)2Hchiral ligand

[Pd(C3H5)Cl]2BSA, KOAc

R' R'

OAcH2C(CO2R)2+

Nu

Pd

ON P

PhPh

Introduction

9

Scheme 1.8

X-Ray structures of the allylic chiral palladium complexes and mechanistic studies ofthis asymmetric allylic substititution have been reported. Attack of the nucleophile42

must oc cur trans to the Pd-P bond, shown in Figure 1.1, furnishing the observe denantiomer.Furthermore , when ligand 1.11 , with a stereogenic phosphorus center (Ar … Ar ), were1 2

employed as catalyst, cycloalkenyl acetates could for the first time be alkylated withrelatively high enantioselectivities (up to 85%). 43

Figure 1.1

Chapter 1

# Throughout thi s thesis a distinction have been made between the conjugate addition reaction of organiccompounds, for instance diethyl malonate (Michael addition) and the conjugate addition reaction oforganometallic reagents (1,4-addition).

10

1.3 Aims of this thesis and survey of its contents

As outlined in the last Section, several examples are known of catalytic asymmetriccarbon-carbon bond formations for a wide variety of substrates, furnishing valuableproducts in high yields and with excellent enantioselectivities. However, the carbon-carbon bond formatio n by conjugate addition of carbon nucleophiles to ",$-unsaturatedcompounds (1,4-addition), being one of the basic reactions in organic synthesis, stillla cks a highly regioselective and enantioselective catalyst with a broad scope. Thi sthesis describes the development of several catalytic systems capable o fenantioselective conjugate addition reactions of organometallic reagents to enones.A number of examples are currently known of both Michael type additions and 1,4-additions of organometallic reagents catalysed by chiral metal complexes wit hen antioselectivities exceeding 80%. In Chapter 2 a comprehensive survey of al l#

catalytic enantioselective conjugate addition reactions is given.Chapter 3 details the investigation on the nickel / chiral amino alcohol catalyse daddition of dialkylzinc reagents to several chalcones. Preliminary experiments werealready performed by Jansen. In Chapter 4 the synthesis of several novel chiral tri-44

and tetradentate ami no alcohol ligands and the results of their examination in the nickelcata lysed conjugate addition of diethylzinc to chalcone and cyclohexenone ar edescribed.Chapter 5 describes the development of a catalytic system capable of enantioselectiveconjugate addition reactions of organometallic reagents to cyclic and acycli csubstrates. Cobalt and copper catalysed conjugate additions of diethylzinc as well asthe copper catalysed addition reactions of trimethylaluminium were investigated.In Cha pter 6 several novel phosphorus amidites, representing a new class of chira lligan ds, are presented. The synthesis and application as chiral ligand in the coppe rcatalysed conjugate addition of dialkylzinc reagents to cyclic and acyclic enones, arereported. The results of the application of representatives of this new class of ligands in severalother catalytic addition reactions are described in Chapter 7.Parts of this thesis have already been published, or will be published.45 46

Introduction

11

1. For the early history of optical isomerism, see Eliel, E.L. Stereochemistry of Carbon Compounds;McGraw-Hill: New York, 1962 .

2. For extensive reviews on stereoselective reactions, see: a) Asymmetric Synthesis; Morrison, J.D. Ed.;Academic Press: Orlando, 1983-1985 , Vol. 1-5. b) Topics in Stereochemistry; Eliel, E.L.; Wilen, S.H.Eds.; Wiley and Sons: New York, 1967-1990 , Vol 1-20.

3. a) Pfeiffer, C.C. Science 1956 , 124, 29. b) Sheldon, R. A. Chirotechnology; Marcel Dekker, Inc., NewYork, 1993 , Chapter 2 and references therein.

4. Merck Index , Budavari, S. ed., 11 ed. Merck 1990 , No. 9182, p. 1458. Recently, the actions of racemicth

and enantiomerically pure thalidomide were reinvestigated. Probably racemisation in vivo occurs, see:Eriksson, T.; B jörkman, S.; Roth, B.; Fyge, Å.; Höglund, P. Chirality 1995 , 7, 44 and references therein.

5. Reference 2a, Vol 1, Chapter 1.

6. Jacques, J.; Collet, A.; Wilen, S.H. Enantiomers Racemates and Resolutions; Wiley: New York, 1981 .See also reference 3b, Chapter 6.

7. Kagan, H.B.; Fiaud, J.C. Topics in Stereochemistry; Eliel, E.L.; Wilen, S.H. Eds.; Wiley and Sons: NewYork, 1988 , 18, 249.

8. a) Wong , C.-H.; Whitesides, G.M. Enzymes in Synthetic Organic Chemistry; Tetrahedron Organi cChemistry S eries, No. 12; Pergamon: Oxford, 1994 . b) Van der Deen, H.; Cuiper, A.D.; Hof, R.P.; VanOeveren, A.; Feringa, B.L.; Kellog, R.M. J. Am. Chem. Soc. 1996 , 118, 3801 and references therein.

9. Hanessian, S. Total Synthesis of Natural products: The 'Chiron' Approach; Pergamon Press: Oxford,1983 . See also reference 3b, Chapter 5.

10. Blaser, H.-U. Chem. Rev. 1992 , 92, 935.

11. See reference 3b, Chapter 4.

12. Reference 8a, Chapter 4.

13. Atkinson, R.S. Stereoselective Synthesis; Wiley: Chichester, 1995 . See also reference 2.

14. a) Catalytic Asymmetric Synthesis; Ojima, I., Ed.; VCH Publishers: New York, 1993 . b) Brunner, H.;Zettlmeier, W. Handbook of Enantioselective Catalysis; VCH Publishers: New York, 1993 . c) Noyori,R. Asymmetric Catalysis in Organic Chemistry; Wiley: New York, 1994 . d) Advances in CatalyticProcesses, Asymmetric Chemical Transformations; Doyle, M.P., Ed.; Jai Press: Connecticut, 1995 .

15. For non-metal based chiral catalysts, see for example Wynberg, H. Topics in Stereochemistry; Eliel,E.L.; Wilen, S.H. Eds.; Wiley and Sons: New York, 1988 , 16, 87.

16. a) Hof, R.P. Enantioselective Synthesis and (Bio)catalysis, Ph.D. Thesis, University of Groningen, 1995 .b) Xianming, H. Chiral Nonracemic 1-Aryl-2,2-dimethyl-1,3-propanediols: Reactions and Applicationsin Asymmetric Synthesis, Ph.D. Thesis, University of Groningen, 1995 . c) Rispens, M.T .Enantioselective Oxidation Using Transition Metal Catalysts, Ph.D. Thesis, University of Groningen,1996 . d) Vries, T.R. Chiral Cyclic Derivatives of C-Symmetrical Butanedioic Acids, Ph.D. Thesis ,2

University of Groningen, 1996 . e) Knol, J. Chiral Lewis Acid Catalyzed Diels-Alder Reactions, Ph.D.Thesis, University of Groningen, 1996 .

17. Nozaki, H.; Moriuti, S.; Takaya, H.; Noyori, R. Tetrahedron Lett. 1966 , 5239.

18. Aratani, T. Pure Appl. Chem. 1985 , 57, 1839 and references therein.

19. Doyle, M.P. Recl. Trav. Chim. Pays-Bas 1991 , 110, 305.

1.4 References and notes

Chapter 1

12

20. a) Fritschi, H.; Leuteneggar, U.; Pfaltz, A. Angew. Chem., Int. Ed. Engl. 1986 , 25, 1005. b) Fritschi, H.;Leuteneggar, U.; Siegmann, K.; Pfaltz, A.; Keller, W.; Kratky, Ch. Helv. Chim. Acta 1988 , 71, 1541.

21. a) Müller, D.; Umbricht, G.; Weber, B.; Pfaltz, A. Helv. Chim. Acta 1991 , 74, 232. b) Lowenthal, R.E.;Abiko, A.; Masamune, S. Tetrahedron Lett. 1990 , 31, 6005. c) Evans, D.A.; Woerpel, K.A.; Hinman,M.M. J. Am. Chem. Soc. 1991 , 113, 726. d) Evans, D.A. Woerpel, K.A.; Scott, M.J. Angew. Chem., Int.Ed. Engl. 1992 , 31, 430.

22. Only leading references are noted. Boron: a) Furut a, K.; Shimizu, S.; Miwa, Y. Yamamoto, H. J.Org. Chem. 1989 , 54, 1481. b) Corey, E.J.;Loh, T.-P. J. Am. Chem. Soc. 1991 , 113, 8966. c) Ishihara, K.; Yamamoto, H. J. Am. Chem. Soc. 1994 ,116, 1561.Aluminium: d) Maruoka, K.; Itoh, T.; Sakurai, M.; Nonoshita, K.; Yamamoto, H. J. Am. Chem. Soc.1988 , 110, 3588. e) Corey, E.J.; Imai, N.; Pikul, S. Tetrahedron Lett. 1991 , 32, 7517. f) Corey, E.J.;Sarshar, S.; Lee, D.-H. J. Am. Chem. Soc. 1994 , 116, 12089. g) Bao, J.; Wulff, W.D. J. Am. Chem. Soc.1993 , 115, 3814.Titanium: h) Narasaka, K.; Inoue, M.; Yamada, T. Chem. Lett. 1986 , 1967. i) Mikami, K.; Motoyama,Y.; Terada, M. J. Am. Chem. Soc. 1994 , 116, 2812. Copper: j) Evans, D.A.; Miller, S.J.; Leckta, T. J. Am. Chem. Soc. 1993 , 115, 6460.Lanthanide: k) Kobayashi, S.; Ishitani, H. J. Am. Chem. Soc. 1994 , 116, 4083.

23. a) Oguni, N.; Omi, T. Tetrahedron Lett. 1984 , 25, 2823. b) Kitamura, M.; Suga, S.; Noyori, R. J. Am.Chem. Soc. 1986, 108, 6071. c) Soai, K.; Ookawa, A.; Kaba, T.; Ogawa, K. J. Am. Chem. Soc. 1987 , 109,7111.

24. a) Noyori , R.; Kitamura, M. Angew. Chem. Int. Ed. Engl. 1991 , 30, 49. b) Soai, K.; Niwa, S. Chem. Rev.1992, 92, 833. c) Knochel, P.; Singer, R.D. Chem. Rev. 1993 , 93, 2117. See also reference 14c, Chapter5 and reference 16a, Chapter 4.

25. For a brief review, see: Bach, K. Angew. Chem., Int. Ed. Engl. 1994 , 33, 417.

26. Yanagisawa, A.; Nakashima, H.; Ishiba, A.; Yamamoto, H. J. Am. Chem. Soc. 1996 , 118, 4723 an dreferences therein.

27. Mikami, K.; Matsukawa, S. J. Am. Chem. Soc. 1993 , 115, 7039.

28. a) Costa, A.L.; Piazza, M.G.; Tagliavini, E.; Trombini, C.; Umani-Ronchi, A. J. Am. Chem. Soc. 1993 ,115, 7001. b) Keck, G.E.; Tarbert, K.H.; Geraci, L.S. J. Am. Chem. Soc. 1993 , 115, 8467.

29. See also: a) Faller, J.W.; Sams, D.W.I.; Liu, X. J. Am. Chem. Soc. 1996 , 118, 1217. b) Sawamura, M.;Sudoh, M.; Ito, Y. J. Am. Chem. Soc. 1996 , 118, 3309. c) Evans, D.A.; Murry, J.A.; Kozlowski, M.C. J.Am. Chem. Soc. 1996 , 118, 5814.

30. Becker, W.; Freund, H.; Pfeil, E. Angew. Chem., Int. Ed. Engl. 1966 , 4, 1079.

31. Tanaka, K.; Mori, A.; Inoue, S. J. Org. Chem. 1990 , 55, 181 and references therein.

32. Reetz, M.T.; Kunisch, F.; Heitmann, P. Tetrahedron Lett. 1986 , 39, 4721.

33. Nitta, H.; Yu, D.; Kudo, M.; Mori, A.; Inoue, S. J. Am. Chem. Soc. 1992 , 114, 7969 and reference stherein.

34. For a mechanistic study on this reaction, see: Shvo, Y.; Gal, M.; Becker, Y.; Elgavi, A. Tetrahedron:Asymmetry 1996 , 7, 911.

35. Denmark, S.E.; Nakajima, N.; Nicaise, O. J.-C. J. Am. Chem. Soc. 1994 , 116, 8797.

36. Iyer, M.S.; Gigstad, K.M.; Namdev, N.D.; Lipton, M. J. Am. Chem. Soc. 1996 , 118, 4910.

Introduction

13

37. Trost, B.M.; Strege, P.E. J. Am. Chem. Soc. 1977 , 99, 1649.

38. Fiaud, J.C.; Hibon de Gournay, A.; Larcheveque, M.; Kagan, H.B. J. Organomet. Chem. 1978 , 154, 175.

39. Reviews: a) Fros t, C.G.; Howarth, J.; Williams, J.M.J. Tetrahedron: Asymmetry 1992 , 4, 1089. b) Reiser,O. Angew. Chem., Int. Ed. Engl. 1993 , 32, 547.

40. Pfaltz, A. Acc. Chem. Res. 1993 , 26, 339 and references therein.

41. a) Von Matt, P.; Pfaltz, A. Angew. Chem., Int. Ed. Engl. 1993 , 32, 566. b) Sprinz, J.; Helmchen, G .Tetrahedron Lett. 1993 , 34, 1769. c) Dawson, G.J.; Frost, C.G.; Williams, J.M.J. Tetrahedron Lett. 1993 ,34, 3149.

42. a) Sprinz, J.; Kiefer, M.; Helmchen, G.; Reggelin, M.; Huttner, G.; Walter, O.; Zsolnai, L. TetrahedronLett. 1994 , 35, 1523. b) Von Matt, P.; Lloyd-Jones, G.C.; Minidis, A.B.E.; Pfaltz, A.; Macko, L. ;Neuburger, M.; Zehnder, M.; Rüegger, H.; Pregosin, P.S. Helv. Chim. Acta 1995 , 78, 265.

43. Sennhenn, P.; Gabler, B.; Helmchen, G. Tetrahedron Lett. 1994 , 35, 8595.

44. Jansen, J.F.G.A.; Feringa, B.L. Tetrahedron: Asymmetry 1992 , 3, 581.

45. Chapter 2 : Feringa, B.L.; de Vries A.H.M. in Advances in Catalytic Processes, Vol 1: AsymmetricChemical Transformations; Doyle, M.P., Ed.; JAI Press, Connecticut, 1995 , p. 151.Chapter 3: de Vries, A.H.M.; Jansen, J.F.G.A.; Feringa, B.L. Tetrahedron 1994, 50, 4479.Chapter 6: de Vries, A.H.M.; Meetsma, A.; Feringa, B.L. Angew. Chem., Int. Ed. Engl. in press.

46. Chap ter 4: de Vries, A.H.M.; Imbos, R.; Feringa, B.L. manuscript in preparation for Tetrahedron:Asymmetry.Section 5.2: de Vries, A.H.M.; Feringa, B.L. manuscript in preparation for Tetrahedron: Asymmetry.Section 5.3: de Vries, A.H.M.; Hof, R.P.; Staal, D.; Kellogg, R.M.; Feringa, B.L. submitted t oTetrahedron: Asymmetry.

(2.1)

EWG = CHO, COR, CO2R, CONR2, CN, SO2R, NO2, etc.

REWG R

EWGNu

E-REWG

NuNu- E+

Nu-

E+Nu-

E+ R R'E

Nu O

R R'

OENu

R R'

O

1,4-addition1,2-addition

(2.2)

1

Chapter 2

Carbon-Carbon Bond Formation byCatalytic Enantioselective Conjugate Addition

2.1 Introduction

Conjugate addit ion reactions of carbon nucleophiles to ",ß-unsaturated compounds aream ong the most widely used methods for carbon-carbon bond formation in organi csynthesis. It is therefore not surprising that major efforts have been devoted to achieve1

asymme tric conjugate addition despite the often complicated nature of many 1,4 -addit ion reactions. Addition of the nucleophile to the ß-position of an electron -2,3

deficient alkene re sults in a stabilized carbanion. After protonation of the carbanion (E +

= H ) a ß-substituted product is formed. Quenching of the stabilized carbanion with+

electrophile s provides ",ß-disubstituted products with two newly created stereocenters(Eq. 2.1).

As carbon nucleophiles one can use a variety of organometallic reagents, "classical"Mi chael donors, carbanions derived from nitro alkanes, nitriles or dithianes, an denolates (and deri vatives). Common substrates for conjugate addition reactions are ",ß-unsaturated ald ehydes, ketones, esters, amides, nitriles, sulfones, and nitro compounds.Typical problems associated with conjugate addition are regioselectivity an drevers ibility. Competition between 1,2- and 1,4-addition to enones is governed b y4

several parameters, but in general the use of soft carbon nucleophiles results in highselectivities for conjugate addition products (1,4-addition, Eq. 2.2).

X

Y

_ EWGR

X Y

_

REWG+ (2.3)

Chapter 2

2

In Michael additions, which are often executed under thermodynamic controlle dconditions employing stabilized carbanions, reversibility is not an uncommon feature(Eq. 2.3). T he stereochemistry might be affected by such reversible processes, whereasthe presenc e of a labile hydrogen at the "-carbon of the product (with respect to EWG)could be another complicating factor as racemisation or epimerisation can occur.

The enormous utility of conjugate addition reactions in synthesis is partly a result ofthe la rge variety of donor and acceptor compounds that can be employed. Anothe rimportant aspect is the high diastereoselectivity often observed. These features havebeen a strong impe tus for the development of enantioselective conjugate additions. Theuse of natural product-based Michael type acceptors has been extremely successful,commonly leading to Michael products with high and predictable stereoselectivities. 2-4, 5

Stoichiometric asymmetric conjugate additions have been developed along two lines:(1) using a chiral auxiliary-based Michael acceptor i.e. chiral ",$-unsaturated ester ,amide, sulfoxide, or (2) by reaction of a chiral reagent with a prochiral electron -2,5, 6

deficient alkene. In the latter case two strategies have been used mainly, namel y2,3,6

chiral auxil iary based donors, such as enamines and enol derivatives, and chiral ligandmo dified organometallic reagents, in particular chiral cuprates, Grignard reagents ,org anozincates, and organolithium reagents. Natural product-based and syntheti corganic ligands and auxiliaries have been successfully employed, but hig hdiaster eoselectivities were reached also with organometallic auxiliaries. Asymmetric7

conjugate addition reactions using stoichiometric chiral auxiliary-based Michae ldo nors and acceptors and chiral organometallic reagents have been extensivel ycovered by reviews and the reader is referred to these papers for specific examples. 1-6

It should be emphasised that several chiral auxiliary-based acyclic and cyclic ",ß-unsatu rated substrates, enolates, and enamines are now available which giv eenantioselectivities exceeding 95% in a variety of reactions. Furthermore, there are anumber of organocopper reagents with chiral non-transferable ligands as well a sorganocuprates modified by additional chiral ligands known today, that provide 1,4-addition products with e.e. > 95%.Ho wever, only for a limited number of prochiral acyclic and cyclic enones hig henantioselectivities are reached ( vide infra). Major improvements are necessary, i n3

R R'

O(R'')mMLn +

R R'

R'' OMLn(2.4)

M = Cu, Li, Zn, Mg

Carbon-Carbon Bond Formation by Catalytic Enantioselective Conjugate Addition

3

parti cular with respect to the scope of chiral reagent-based methods. This become sevident when one co nsiders applications in practical synthesis of enantiomerically purecompounds employing conjugate addition as a key step.Even more challengi ng is the development of general methodology for enantioselectivecarbon-carbon bond formation using chiral non-racemic catalysts in combination withreadily available organometallic reagents and Michael donors. A literature survey ofthis area i s the subject of this Chapter with the emphasis on enantioselective conjugateaddition catalysed by chiral transition metal complexes.

2.2 Asymmetric metal-mediated 1,4-addition

In ord er to achieve a rational synthesis of new chiral catalysts for enantioselectiv econjugate add ition it is important to consider several factors that might govern the 1,4-addition step. Among these are: (1) the nature of organometallic reagent (R'') M (Eq.m

2.4), (2) the ligands L associated with it, (3) the fact that most of these reagents aren

aggregated in s olution (solvent dependent), and (4) the notion that stereoselectivity (aswell as regioselectivity) can be affected by additional ligands, coordinating solventsand salts.

Furthe rmore, activation of the electron deficient alkene by Lewis acid or catio ncomplexation to the carbonyl moiety is often proposed as a means to tether the reagent,catalyst, and enone i n order to increase stereoselectivity and enhance reactivity towardswe aker nucleophiles. The coordinating metal can either be from the organometalli creagent, the catalyst, or additional metal ions ( i.e. salts). The proposed intermediate Iin the highly enantioselective (90% e.e.) conjugate addition of the (1 R,2S)-ephedrine-based mixed cuprate, reported by Corey and co-workers, nicely illustrates additional8

lith ium ion coordination between the oxygen of the enone and the cuprate ligan d(Figure 2.1).

O Li

Cu

RLi

N

Ph

NO

Ln

II (Ln = I- or THF)

IIII

R R'

OLA

LA = Me3SiCl, BF3

+ (R'')mMLnR R'

R'' OLA

OO O

O+ O

O

IVaIVa IVbIVbIIIIII 99 % 1 %

(2.5b)

(2.5a)

Me2CuLiMe3SiCl-78°C, 2 min

Chapter 2

4

Figure 2.1 Proposed intermediate in the enantioselective conjugate addition of amixed cuprate reported by Corey and co-workers.8

The use of Lewis acid s, in particular Me SiCl or BF , often results in a dramatic increase3 3

of reacti on rates in 1,4-addition reactions of cuprates, presumably by enone activation(Eq. 2.5a). Increased stereoselectivity, for instance, almost exclusive formation of IVa9

by Me CuLi addition to III in the presence of Me SiCl (Eq. 2.5b), and the formation2 310

of enol de rivatives such as II, which can subsequently be used in electrophili cadditions (tandem 1,4 -addition-enolate processes), are additional important advantages.

Lewis acid catalysis has been extremely successful in 1,4-additions of enol silyl ethers(and ti n-analogues). The role of the Lewis acid can be an activation of the enone and11

the silyl-enolat e leading, via a cyclic transition state V (Figure 2.2), to Michael adductswith high stereoselectivities.

OSi X

MLnO

RR''

R'

VV (XMLn = Lewis acid)

X

H

R'

Y

LnM

B

VIIVII (B = base)VIbVIb (Z-enolate)VIaVIa (E-enolate)

OR

X

LnMO

X

MLn

R

(2.6)

R = i-Pr ; R' = OMe

2 : 98

93 : 7

R = Me ; R' = Ph

+ t-BuR'

R OO

t-BuR'

O R O

O

R'

Si+

R t-Bu

O TiCl4


5

Figure 2.2 Activation of enone and silyl enolate by a Lewis acid.

The h igh level of regulation in V that might be reached in these cases offers attractivepossibilities for the development of new chiral Lewis acid catalysts for 1,4-addition.Moreo ver, stereoselectivity appears to be strongly Lewis acid- and substituent -dependent, as is illustrated in Eq. 2.6. 12

Figure 2.3

When enolate anions or other stabilized carbon nucleophiles are involved in conjugateaddition reactions in the presence of chiral metal catalysts, the catalyst can exert its

R''O H

H R'LnM

O R

X H

R'H

HOR''

R

XH

OLnM

(2.7a)

(2.7b)

anti

syn

R''

O

X

O

R

R'

R''

O

X

O

R

R'

Chapter 2

6

stereodirecting effect via formation of a chiral metalenolate VI (Figure 2.3) or b ycomplexation and activation of the Michael donor VII (or both donor and acceptor).The geometry of the enolate ( VIa or VIb ) is a decisive factor in the B-face selectivityand syn-anti diastereoselectivity. Finally, it should be noted that the stereochemical13

result o f 1,4-additions of enolate type carbon nucleophiles strongly depends o nchelation or non-chelation control. 6,13

In Eq. 2.7a and Eq. 2.7b the open transition state model and the closed (chelated )transition state model for metalenolate additions to enones, leading to syn- and anti-adducts, respectively, are given for the case of a Z-enolate.

A number of highly diastereoselective conjugate additions of metalenolates has beendeveloped, strongly stimulated by the results of stereoselective aldol reactions using6,14

well defined E- and Z-metalenolates and by exploring the coordination properties ofseveral metals (in particular B, Li, Ti and Mg). Again enolate geometry, solvent ,counterion, metal catalyst, and mode of addition can have a strong influence on thester eochemical result and all these variables need to be considered in designing a neffective chiral catalyst for 1,4-addition.The use of chiral metal catalysts for enantioselective carbon-carbon bond formationusing organometallic reagents and other carbon nucleophiles will be described in thenext sections. For the sake of completeness conjugate addition reactions using chiralcrown ethers and bases are included.

CuBr Me2S (cat.)2.14 or 2.15 (cat.)

.HMPA, R3SiClTHF, -78°C

14-74% e.e.

RMgX+

2.2

(2.8)R

OO


7

2.3 Enantioselective conjugate addition of Grignard reagents

Chiral copper complexes as catalystsRapid pr ogress has been made in recent years towards highly enantioselectiv econj ugate addition reactions of chiral, ligand modified, cuprates and organocoppe rre agents with non-transferable ligands based on Grignard and organolithiu mcompounds. It is surprising that despite many attempts, only very recently the firs t3

examples of successful 1,4-additions of Grignard reagents catalysed by chiral coppercomplexes were reported, albeit modest selectivities and limited scope were reached.Lippard and co-workers described the first catalytic conjugate addition of n-BuMgBrto 2-cyclohexen -1-one ( 2.2) (Eq. 2.8). Using a copper(I) complex derived in situ from15

chir al N,N'-dialkylaminotropone imine (H-( R)-CHIRAMT, 2.14 , Figure 2.5), 3 -butylcy clohexanone was obtained in 14% e.e. The enantioselectivity of the reaction issignificantly i ncreased by the addition of hexamethylphosphoric triamide (HMPA) andsilyl reagents (see Section 2.2).

Both HMPA and a bulky silyl reagent seem to be essential to reach high e.e.'s. Th ehighest enantioselectivity (e.e. 74%, yield 53-57%) is obtained using 2 equivalents oft-butyldiphenylsi lylchloride, 2 equivalents of HMPA and 4 mol% of chiral ligands 2.14or 2.15 . The role of HMPA and the silyl reagent remains rather obscure at present, butit seems that these additives suppress the uncatalysed conjugate addition. A slightly16

higher e.e. (78 %) is obtained when a stoichiometric amount of copper and chiral ligandis used. Compared to n-BuMgCl poor enantioselectivities were found with MeMgCl orEtMgCl (30 and 14% e.e. respectively). Interestingly, the reaction with MeMgCl gave(R)-3-methylcyclohexanone in excess, instead of the S enantiomer obtained in th ereaction with n-BuMgCl. This reversal indicates that the Grignard reagent is involvedin the rate d etermining step of the reaction. Other enones are converted as well, thoughno enantioselectivity was observed. All these effects clearly demonstrate the complexnature of the catalytic sequence and in particular further study of additive effects willbe necessary. Since the pioneering work of Lippard and co-workers several othe rgr oups have reported catalytic enantioselective conjugate additions of Grignar d

2.1 n = 0 ; X = CH 22.2 n = 1 ; X = CH 22.3 n = 2 ; X = CH 22.4 n = 1 ; X = O

X

O

( )n

R R'

O

2.5 R, R' = Ph 2.6 R = Ph ; R' = Me 2.7 R = Me ; R' = Ph 2.8 R = Ph ; R' = CPh 3 2.9 R = Ph ; R' = t-Bu2.10 R = Ph ; R' = p-MeOPh2.11 R = p-MeOPh ; R' = Ph2.12 R = p-ClPh ; R' = Ph2.13 R = Ph ; R' = p-ClPh

(R' = i-Pr, t-Bu ; X = Cl, Me, OMe)

MeMgI+

2.6 R' = Me ; X = H

R'

O

X

2.16 (cat.)

Et2O, 0°C(2.9)

45-76% e.e.

R'

O

X

Chapter 2

8

reagents. The relev ant results will be described in this section. The substrates and chiralcatalysts are compiled in Figure 2.4 and 2.5, respectively.

Figure 2.4 Substrates used in catalytic enantioselective conjugate additionreactions of Grignard and dialkylzinc reagents (see sections 2.3 and 2.5).

Van Koten and co-workers have reported the use of chiral copper(I) arenethiolate (2-[1-( R)-(dimethylamino)ethyl]phenylthiolate copper(I), ( 2.16 )) as a catalyst for th eenantio selective addition of MeMgI to benzylideneacetone ( 2.6) (Eq. 2.9). The17

ena ntioselectivity is highly dependent on the mode of addition. Only controlled ,simultaneous addition of solutions of MeMgI and of 2.6 (at equal concentration) t oca talyst 2.16 (9 mol%) in diethyl ether afforded a relatively high enantioselectivit y(76% e.e.). This indicates that a cuprate reagent rather than free MeMgI is involved inthe reaction.

Using the optimal parameters found for MeMgI, the scope of this reaction has bee nexamined for other Grignard reagents ( n-BuMgI and i-PrMgI, 45 and 10% e.e. ,respectiv ely) and various acyclic enones (Eq. 2.9). Substrates with different par a

2.22

N N PhPh

ZnCl Ot-Bu

NN

OH

OHOHN

PPh2

Me2N O

N NOH

2.25

2.26 2.27 2.28

OH

OH

NH2

NOH

H

i-Pr

SH N

O

R

2.17 R = Me2.18 R = i-Pr2.19 R = t-Bu2.20 R = benzyl

2.16 (trimer)

2.24 2.23

2.21

OSH

OO

O

ON

N

R

R

H

2.14 R = Ph2.15 R = 1'-naphthyl

N

S Cu


9

sub stituents on the aromatic ring and steric bulk next of the carbonyl group gav eproducts with slightly lower enantioselectivities (45-72% e.e.). However, the use o fchalcone ( 2.5) (or cyclohexenone) furnished the 1,4-product with an e.e. of 0%.

Figure 2.5 Chiral ligands and complexes used as catalysts in the conjugate additionof Grignard reagents to enones.

Copper(I) thiolate complexes derived in situ from chiral mercaptophenyloxazoline s2.17-2.20 , were reported by Zhou and Pfaltz to be effective in the 1,4-addition of n-BuMgCl to cyclic e nones 2.1-2.3 (Eq. 2.10). Highest enantioselectivities were reached18

only when the Grignard reagent was added slowly at -78 EC to the solution of enone,catalyst, and two equivalents of HMPA. The methyl and i-propyl derivatives 2.17 and2.18 were found to be the most effective ligands, whereas the bulky derivative 2.19gave markedly lower e.e.'s. Significant enantioselectivities were found only in th epresen ce of HMPA. The use of trialkylchlorosilanes as additives ( vide supra) resulted

O

( )n

O

R( )n

CuI (cat.)2.18 (cat.)

HMPATHF, -78°C

16-87% e.e.

RMgX+

2.1-2.3

(2.10)

70-92% e.e.2.2-2.4

(2.11)X

O

( )n

X

R

O

( )nn-alkylMgCl+

CuI (cat.)2.22 (cat.)

Et2O, -78°C

Chapter 2

10

in substantial loss of selectivity. With ligand 2.18 the enantioselectivity of the 1,4 -prod uct increased with ring size of the cyclic enone (cyclopentenone, 16-37% e.e. ;cy clohexenone, 60-72% e.e.; cycloheptenone, 83-87% e.e.). With respect to th e18

Grig nard reagent it was found that i-PrMgCl gave consistingly higher e.e.'s than n-BuMgCl, whereas PhMgBr gave virtually racemic products. Despite the strong analogybetween chiral catalysts 2.16 and the copper catalyst derived from ligands 2.17-2.20 ,only low enantioselectivities (e.e. < 20%) with acyclic enones were reported in th elatter case.

A third catal ytic system based on chiral copper(I) thiolate complexes was described bySpescha and Rihs. The chiral complex, prepared in situ from the lithium salt o f19

1,2,5,6-di- O-isopro pylidene-3-thio- "-O-glucofuranose ( 2.21 ) and CuI, gave high yieldsand regioselectivities exceeding 98% in the addition of n-BuMgCl to enone 2.2 . Theenantioselectivity is strongly dependent on the reaction conditions and variation of alarge numb er of reaction parameters resulted in e.e.'s up to 60%. The catalytic reactionwa s carried out by slow simultaneous addition of a solution of n-butylmagnesiu mhalide and a solution of enone to a solution of the copper complex in order to avoidexcess of reagents . A remarkable dependency on the halide in the Grignard reagent wasobserved and with PhMgBr only an e.e. of 20% was found. Reproducible results werefound upon addition of a radical scavenger (2,2,6,6-tetramethylpiperidin- N-oxyl ,TEMPO) . Together with the dependency of the enantioselectivity on the turnove rnumbers, salts, and solvents these findings typically illustrate the complex nature ofthese catalytic systems.Recently, Tomioka and Kanai reported the use of a chiral bidentate phosphine ligandin the copper catalysed addition of Grignard reagents to cyclic enones. The addition20

of several n-alkylmagnesium chlorides to enones 2.2 or 2.3 in the presence of 8 mol%of CuI and 32 mol% of 2.22 gave 1,4-products in good yields and with relatively highenantioselectivities (70-92% e.e., Eq. 2.11).

(2.12)

2.2

+ i-PrMgX

15-33% e.e.

O

i-Pr

O

THF, -90°C

ZnCl2 (cat.)2.24-2.28 (cat.)


11

Grig nard reagents, such as Me-, Ph-, and i-Pr-magnesium chloride and Grignar dreagents prepared from the corresponding bromides or iodides, gave low e.e.'s and pooryields. Remarkably, the addition of n-butyl- or n-hexylmagnesium chloride to 5,6 -dihydro-2 H-pyran-2-one ( 2.4) was also effective, furnishing synthetically interestingchiral intermediates (90% e.e.).Without doubt, the copper catalysts are highly promising in enantioselective conjugateadditions of Grignard reagents and the first examples of e.e.'s exceeding 95% seem tobe in clos e reach. One of the major problems to deal with, besides tuning of ligand andreaction conditions, are the different aggregates of the catalyst which are in equilibriumwith each other, as they apparently are formed in the reaction medium. The aggregateformation probably d epends on the concentration of reactants and additives, and resultsin different catalytically active species with different enantioselectivities.

Chiral zinc complexes as catalystsThe development o f chiral zinc(II) complexes as catalyst for 1,4-addition reactions wasbase d on the discovery of Isobe and co-workers of the facile conjugate addition o flit hium triorganozincates. Subsequent studies resulted in selective alkyl grou p21

transfer from mixed trialkylzincates, the use of alkoxides as non-transferable ligands,22

and 1,4-additions of Grignard reagents mediated by N,N,N',N'-tetramethylenediaminezinc dichloride as reported by Jansen of our research group. 23

Stimulate d by these results stoichiometric and catalytic amounts of chiral diaminezinccomplex 2.23 (Figure 2.5) were used in the 1,4-addition of i-PrMgBr to cyclohexenone(2.2 ). A catalytic amount (1 mol %) of 2.23 substantially increases yields ,regioselectivities to wards 1,4-adducts and enantioselectivities of the conjugate addition(21% e.e.). A numb er of chiral catalysts were prepared in situ from chiral ligands 2.24-24

2.28 and ZnCl and screened in the model reaction (Eq. 2.12).225

In all cases yield s and regioselectivities are excellent and the highest enantioselectivity(33%) was found with 5 mol% of chiral ligand 2.28 and i-PrMgBr as Grignard reagent.

N N

OZn

C6H5

MgR

X

R

O

Chapter 2

12

The enantioselectivity depends furthermore on a number of variables: 25

- With other alkyl- and aryl-Grignard reagents lower e.e.'s, as compared to i-PrMgX, were found.

- Both regio- and enantioselectivity improved by decreasing the temperature.- The effect of chloride or bromide in RMgX on the enantioselectivity reverses

wit h different chiral ligands. Similar effects have been observed in cuprat eadditions. 19

- A signi ficant improvement of the enantioselectivity due to the presence o flithium ions was ob served and the catalyst has preferentially to be prepared fromthe lithium salt of the ligand (for the lithium ion effect, see also Section 2.2).

- Highe r enantioselectivities were attained by slow addition of i-PrMgX to asolution containing substrate and catalyst. It appears to be essential to keep alow concentration of organometallic species to prevent uncatalysed addition.

The modest ena ntioselectivities and the sensitivity to a large number of variables makeit difficult to po stulate a catalytic cycle. However, the enantioselective 1,4-addition canbe rationalized by a model shown in Figure 2.6, given for the catalyst based on ligand2.28 .

Figure 2.6 Proposed intermediate in the zinc catalysed conjugate addition ofGrignard reagents to 2.2.

Binding of the Grignard reagent via coordination of magnesium to the alkoxide exo tothe bicyclic zinc complex can take place. Activation of the substrate, via coordinationto zinc , involves a pentacoordinated zinc(II) intermediate, which brings Grignar dreagent and enone in close proximity to allow alkyl transfer. In this stage a third metal(i.e. lithium) c ould be involved as proposed for cuprate additions (see also section 2.2).It should be noted that scrambling of both alkyl groups has been observed.The zinc catalysed 1,4-addition of Grignard reagents is attractive as high yields andregios electivities are found although it is obvious that the enantioselectivity need s

(36 mol%)

(2.13)

HONH

NR*OH =

2.29(R)-(-)-muscone85% yield99% e.e.

O

(CH2)6

O

(CH2)6

MeLi, THFtoluene, -78°C

[R*OCuMe2Li2]

2.30


13

substantial improvement.

2.4 Catalytic enantioselective conjugate addition of organolithium reagents

The high rea ctivity commonly found for organolithium reagents compared to Grignardreagents and the preference for 1,2-addition make the development of an efficien tcataly st for conjugate addition of RLi a particularly challenging goal. Significan tenanti oselectivities in catalytic alkyllithium additions have not been reported unti lTanaka a nd co-workers recently realized the chiral alkoxycuprate catalysed addition26

of MeLi to ( E)-cyclopentadec-2-en-1-one ( 2.29 ) affording (R)-(-)-muscone with e.e.99% (Eq. 2.13).

The chiral catal yst was prepared from amino alcohol ligand 2.30 by sequential additionof MeLi, CuI, and MeLi. The conditions for the catalyst preparation are very critical toreach high enant ioselectivities. The use of 1 equivalent of THF, presumably as externalligand to the chiral cuprate, increases the e.e. significantly. Under optimised conditions,36 mol% of chiral ligand 2.30 provides muscone virtually enantiomerically pure andin high yield. Despite impressive e.e.'s in this case further implementation await seffective catalysis at lower catalyst concentration and high selectivities with othe renones.The enantioselective addition of phenyl- and 1-naphthyllithium to 1- and 2 -naphthalenecarboxylic esters of 2,6-di- t-butyl-4-methoxyphenol (BHA) catalysed bychiral diether 2.32 was reported by Tomioka and co-workers (Eq. 2.14). 27

O OBHA CH2OHR

toluene- 45°C

2.32 (cat.)RLi

i. LiEt3BHii. MeOHiii. NaBH4

LiO OBHA

R

2.31

2.32

MeO OMe

Ph Ph 67-75 % e.e.

(2.14)

BA

Me

Zn

Me

N

N145°

1.98 Å

Me2Zn

2N N

N

180°

1.95 Å

Me Zn Me

Chapter 2

14

This is an interesting case of ligand-accelerated organometallic carbon-carbon bondformation. The 1,4-addition in the absence of chiral ligand 2.32 was sluggish. Both28

1- an d 2-hydroxymethyl substituted dihydronaphthalene derivatives have bee nobtained via this catalytic process.

2.5 Conjugate addition of dialkylzinc reagents catalysed by chiral nickelcomplexes

Enantioselective c arbon-carbon bond formation by 1,2-addition of organozinc reagentsto aldehydes has become one of the most successful and active area's of asymmetricsynthesis in recen t years. Although dialkylzinc reagents react extremely sluggish with29

carbonyl compounds, effective catalysis has been achieved by several ligands an dtransition m etal complexes. The catalytic effect was explained by changes in geometryand bond energy of t he zinc reagents. For example, dimethylzinc has a linear structure30

and is not reactive towards aldehydes or ketones (Figure 2.7). Upon coordination oftriazine a tetra hedral configuration at the zinc atom and an elongated zinc-carbon bondis found, resulting in enhanced reactivity of the dialkylzinc reagent.

Figure 2.7 Structures of dimethylzinc (A) and its adduct with 1,3,5-

R R'

O

R R'

O(2.15)

2.5-2.13 43-95% e.e.

Et2Zn+

Ni(acac)2 (cat.)2.33-2.43 (cat.)

CH3CN, -30°C


15

trimethylhexahydro-1,3,5-triazine (B).

Several catalytic 1,4-additions of diethylzinc to acyclic enones employing chiral nickelcomplexes have been developed. The substrates and chiral catalysts are compiled inFigure 2.4 and Figure 2.8, respectively. Based on work of Luche and Greene and co-worker s, an enantioselective modification of the nickel catalysed alkyl transfer from31

diethylzinc to chalcone ( 2.5) was found by Soai and co-workers. The chiral catalyst,32

prepared in situ from NiBr and (1 S,2R)-N,N-di-n-butylnorephedrine ( 2.33 ), afforded2

(R)-1,3-diphenylpentan-1-one in 32% yield with 48% enantiomeric excess. Highe ryields (> 70%) w ere achieved with Ni(acac) instead of NiBr , although large amounts2 2

33

of chiral ligand are required. A remarkable achiral ligand effect was observed .Preparation of the chiral catalyst from 6 mol% of Ni(acac) , 14 mol% of chiral ligand2

2.33 or 2.34 , and 7 mol% of 2,2'-bipyridine in acetonitrile raised the enantioselectivityup to 90%. 34

2.35 2.36

NOH

2.37 2.38

OHSO

PhN

Et

2.39 R = t-Bu2.40 R = (CH2)3Si_{Sup}

2.33 2.34

N(n-Bu)2

Ph Me

HON N

t-Bu t-BuOH HO

Nt-Bu

OH

Ph Me

HO N

NH

O

NHR

OONi

+ PF6

_

NOH

Ph

Cr(CO)3

2.41

NNPhH

n-pentyl

2.42

Ph OBn

OH

N

O 2.43

Chapter 2

16

Co mparable yields and enantioselectivities have been reached with nickel catalyst sprepared in situ from C -symmetric bipyridine 2.35 and chiral pyridine 2.36 , as2

35 36

reported by Bolm and co-workers and amino alcohol 2.37 as found by Jansen in ourlaboratory. 37

The nickel catalysed enantioselective conjugate addition of diethylzinc to chalcone wasalso performed using optically active ß-hydroxysulfoximines as chiral ligands. The38

ligand structure was optimised and an e.e. of up to 70% was reached with ligand 2.38 .

Figure 2.8 Chiral ligands and complexes used as catalysts in the conjugate additionof diethylzinc to acyclic enones.

Sánchez and co-workers reported the conjugate addition of diethylzinc to enones byhomogen eous and supported cationic chiral nickel complexes 2.39 and 2.40 , based onproline ami de ligands. Under homogeneous conditions e.e.'s of 75-77% were reached39

using 5 mol% of catalyst 2.39 at -10 EC. Although the addition reactions were slowerwith the supported chiral complexes 2.40 , the enantioselectivities were raised to 95%.The relativ ely high enantioselectivities observed with a chiral ligand-to-nickel ratio of1, compared to ratios of more than 2 in other studies, are explained by the fact32-38,40-42

+ Et2Zn CuI (cat.)2.44 (cat.)

toluene20°C

2.2

NP

OPh

N

i-Pr2.44

70 % yield32 % e.e. (S)

(2.17)

O

Et

O


17

that a sin gle chiral complex is used. Competing catalysis by achiral Ni(acac) (or other2

complexe s, see: Eq. 2.16) presumably cannot take place, as often happens in othe rcatalytic reactions. An attractive feature of this system is also the easy removal andrecovery of the chiral catalyst.

Ni(acac) + 2 L W Ni(acac)L + L W NiL (2.16)2 2* * * *

1,2-Disubst ituted arene-chromium complex 2.41 was also employed as chiral ligand inth e nickel catalysed 1,4-addition. Modest e.e.'s were found strongly depending o n40

amount of catalyst and structure of chromium complex. Recently, two other examplesof the nick el catalysed asymmetric conjugate addition of diethylzinc to acyclic enoneswere reported. With diamine 2.42 or amino alcohol 2.43 diethylzinc was added to41 42

acycli c enones furnishing the 1,4-products in good yields and with e.e.'s ranging from58-89%.All reports reveal the following observations:- The presence of acetonitrile (or another nitrile) as solvent, and presumably as

stabilizing ligand to nickel, appears to be essential in all cases. - Nickel acetylacetonate was found to be the nickel source of choice.- The enantioselectivity strongly depends on the ligand-to-nickel ratio and th e

concentration of the in situ prepared chiral catalyst.- A l imited number of alkylzinc reagents and substrates has been successfull y

used so far in the 1,4-addition reactions described.. Various acyclic enones ( 2.5-2.13 , Figure 2.4) give high e.e.'s., however, 2-cyclohexenone ( 2.2) and "-ß-unsaturated esters gave racemic products and low yields. 36

Soai and co-workers reported that enantioselective conjugate addition to enones alsoproceed with chiral amino alcohol as catalyst without the use of transition metals ,although at much lower rates. After 4 days of reaction time, 1,4-products with e.e.'s43

of 70-80% were obtained using 25 mol% of 2.34 .Rec ently, Alexakis and co-workers reported the first example of copper catalyse denantioselective c onjugate addition of diethylzinc to 2-cyclohexenone (Eq. 2.17). The44

use of 10 mol% of CuI and 20 mol% of trivalent phosphorous ligand 2.44 resulted in

2.582.45

(2.18)O

O O

OMe

O

+

O O

OMe2.63-2.68 (cat.)

[Co(acac)2 (cat.)]

Chapter 2

18

an en antioselectivity of 32%. Under the same conditions chalcone gave racemi cmaterial.

All these efforts represent a significant advance in the field of catalyti cenantios elective conjugate addition reactions, however, there is no general solution tothe problem of achieving efficient catalysis for a wide variety of enones.

2.6 Catalytic Michael additions

Chiral metal complexes as catalystsCarbon-carbon bond formation via Michael additions are most frequently performedunder conditions of base catalysis. The conjugate addition of 1,3-dicarbony lcompounds to en ones can also be efficiently catalysed by metal complexes. Among theadvantage s of transition-metal catalysed Michael additions are the high yields that areoften found under mild reaction conditions, whereas side reactions, frequentl yen countered in base catalysed Michael additions, are avoided. Several catalyti cMichael additions employing chiral metal complexes have been developed. Th eMichael donors and acceptors and chiral catalysts are compiled in Figures 2.9 and 2.10,respectively.Br unner and Hammer were the first to report significant enantioselectivity in atransition-metal catalysed Michael addition. The addition of methyl-1-oxo-2 -45

indanecar boxylate (2.45 ) to methyl vinyl ketone (MVK, 2.58 ) in the presence of 3mol% of a chiral cobalt(II) complex, derived in situ from Co(acac) and (1 S,1S)-(-)-1,2-2

diphenylethylenediamine (2.63 ), provided the Michael product with a nenantioselectivity of 66% (Eq. 2.18).

In further investigations, the Co(acac) -(-)-1,2-diphenylethylenediamine catalyst was2


19

examined in the Michael addition of unsymmetrical 1,3-dicarbonyl donors unde rvarious conditions. Using MVK, di- t-butyl methylenemalonate ( 2.62 ), and acrolein46

(2.59) as Michael acceptors and Michael donors 2.46 and 2.48 , enantioselectivities upto 37% were reached. Under the reaction conditions the enantioselectivity was almosttemperature independent, no racemisation took place and the conjugate addition wasirreversible.Desimoni and co-workers also investigated the model reaction given in Eq. 2.18 ,employing chir al copper(II) complexes 2.64-2.67 . All copper complexes are based on47

Schiff base ligands derived from salicylaldehyde and chiral amino alcohols and arepresuma bly dimeric structures. Furthermore there is evidence that H O is bound to the2

copper atom in th ese complexes resulting in six coordination around each copper atom.The enantioselectivity strongly depends on the solvent and the chiral catalyst. Withcatalyst 2.64 enantioselectivities up to 54% were found in CCl . A negative facto r4

seems to be the ability of the solvent to compete with the chiral ligand for meta lcomplexation. Introd uction of a phenyl substituent in the chiral catalyst structure ( 2.65 )drasti cally reduces the e.e. (7%). Increase of the rigidity of the catalyst b yincorporating an additional hydroxyl group that can act as an axial ligand in 2.66 and2.67 raised the e.e. to 70% (based on optical rotations). Nearly quantitative yield s47

were observed with 1-10 mol% of copper(II) catalyst at -20 EC in CCl .4

O O

OMe

2.45

OR

O O

2.46 R = Et2.47 R = Bn

O O

OEt

2.48

NCOR

O

2.49 R = Me2.50 R = Et2.51 R = i-Pr2.52 R = t-Bu

R R'

O

RO OR

O O

R'

2.53 R = Me ; R' = H2.54 R = i-Pr ; R' = H2.55 R = t-Bu ; R' = H2.56 R = Bn ; R' = H2.57 R = Bn ; R' = Me

t-BuO Ot-Bu

O O

2.622.58 R = H ; R' = Me2.59 R = H ; R' = H2.60 R = Me ; R' = H2.61 R = Me ; R' = Me

Michael acceptors :

Michael donors :

Chapter 2

20

Figure 2.9 Michael donors and acceptors used in enantioselective Michaeladditions.

An in situ prepared chiral cobalt(II) catalyst derived from Co(acac) and diamino diol2

ligand 2.68 was als o tested in the addition of 2.45 to MVK (Eq. 2.18). With 4-14 mol%48

of 2.68 , acceptabl e yields (46-81%) but low e.e.'s (3-38%) were found. This asymmetriccatalysis is a property of the metal ligand complex and not of the ligand itself (the freeligand 2.68 appears to favour the opposite absolute stereochemistry). With Ni(acac) ,2instead of Co(acac) , a lower enantioselectivity was found.2

48

Recently, Ito and co-workers reported a rhodium catalysed enantioselective Michaeladdition of "-cyanocarboxylates 2.49-2.52 to Michael acceptor 2.58 (Eq. 2.19). The49

chiral catalyst was prepared in situ from RhH(CO)(PPh ) and the trans chelating chiral3 3

diphosphine ligand 2,2''-bis[1-(diphenylphosphino)ethyl]-1,1''-biferrocene (TRAP ,2.69 ). Enantioselectivities ranging from 72-84% (for i-propyl- "-cyanocarboxylate )were observed.

H2NNH2

OH

OH

2.682.64 R = Et2.65 R = Ph2.66 R = (CH2)3OH2.67 R = CH2OH

2.69 TRAP

NCOORb

H

2.70

2.63

H2N NH2

PhPh

2

NO

O

R

Cu

Fe Fe

PPh2

Ph2P

NH

N+Me3

OH-

2.71

NCOR

O+

O

R'

2.49 - 2.52 2.58 or 2.59

2.69 (cat.)RhH(CO)(PPh)3 (cat.)

benzene, 3°C RO R'

OO

NC(2.19)


21

Figure 2.10 Chiral ligands and complexes used as catalysts in enantioselectiveMichael additions.

The reactions of 2.51 with a large variety of vinyl ketones or acrolein ( 2.59 ) proceedwith 83-89% e.e. High catalyst efficiency was observed even with 0.1 mol% of 2.69(84% e.e.) wherea s high yields are generally found. Trans chelation of the chiral ligandto rhodium appears t o be essential for high e.e.'s as common cis chelating diphosphines,such as BINAP , DIOP, or Chiraphos, resulted in low enantioselectivities. It is proposedthat the activated cy anoacetic ester is bound to rhodium through the cyano nitrogen andthat in the enolate intermediate the enantioselective carbon-carbon bond formatio noccurs at the carbon atom rather distant from the metal center (Figure 2.11). Only aconcav e chiral ligand such as TRAP would effect the remote enantiofacia ldif ferentiation. The X-ray crystal structure of trans-[RhCl(CO){( R,R)-(S,S)-n-Bu TRAP}], which bears a n-Bu group instead of a Ph group as in 2.69 , reveals tha trhodium has a nearly planar coordination geometry. The conformation of the ligand50

is essentially C-symmetric, and the chloro and carbonyl groups on rhodium, which may2

be repl aced by a prochiral substrate in a catalytic asymmetric reaction, are completelyburied in the chiral cavity created by the ferrocenyl backbone and the n-Bu groups.

(smaller)

(larger)(TRAP)Rh

OR'CN

-O

electrophile (R)-product

(S)-productelectrophile

Chapter 2

22

Figure 2.11 Proposed mechanism in the rhodium catalysed enantioselective Michaeladdition of "-cyanocarboxylates to vinyl ketones.

Yamaguchi a nd co-workers reported the first catalytic asymmetric Michael addition ofa sim ple malonate ion to prochiral enones and enals. Asymmetric induction wa s51

observed when the Michael addition of dimethyl malonate ( 2.53 ) to prochiral acceptorscatalysed by the lithium salt of L-proline, was carried out in chloroform. Highe rcatalytic activity and enantioselectivity was attained with the rubidium salt 2.70 .Enantioselectiviti es up to 88% were achieved with 5 mol% of 2.70 , malonates ( 2.54 and2.55), and various Michael acceptors such as aliphatic and aromatic enones ( 2.6, 2.60and 2.61), cyclic enones 2.2 and 2.3 , and acrolein. A small amount of water was foundto prom ote the reaction. Yields of Michael products were very low with catalyti camounts of the rubidium salt of N-methyl-L-proline or free L-proline. Thus, both thesecondary amine mo iety and the metal carboxylate moiety of 2.70 are essential for highcatalytic activities. Reversible iminium salt formation to provide chiral Michae lacceptors could account for the above asymmetric inductions (Figure 2.12) .Independent experimen ts demonstrated a high reactivity of an unsaturated iminium salt,derived of 2.6 and pyrrolidine, towards malonate addition. 51

2.6, 2.59-2.61 2.2 or 2.3N

COO-

NR2'' =

R'

NR2''

R

+

-CH(CO2i-Pr)2

( )n

NR2' '+

-CH(CO2i-Pr)2

(2.20)2.25

THF, 0°C

R

O O

R'R''

THF, 0°C

2.72

R

R''

O

OR' La

OO+La(O-i-Pr)3


23

Figure 2.12 Proposed mechanism in the rubidium catalysed enantioselective Michaeladdition.

This reversible iminium salt formation, creating differentiation of enantiofaces of theprochir al Michael acceptor has also been applied successfully with ammoniu mhy droxide 2.71 (Figure 2.10), easily prepared from ( S)-proline. The reaction o f52

malonate 2.53 or 2.56 with cyclic enone 2.1 or 2.2 was conducted in the presence of1,1,1,3,3,3-hexafluoro-2-propanol to reduce the basicity of the catalyst. The additionproducts were isolated in moderate yields (50-60%) and with enantioselectivitie sranging from 56-71% e.e. Rece ntly, Shibasaki and co-workers reported a chiral lanthanum complex, which i shighly effective as catalyst in enantioselective Michael additions of malonates to cyclicenone s. Investigations in order to create an effective catalyst revealed that th e53

li thium-free complex 2.72 , derived from 2,2'-dihydroxy-1,1'-binaphthyl (BINOL )(2.25), La(O- i-Pr) , and a dialkyl malonate furnishes the corresponding products in high3

yield s and e.e.'s ranging from 75-95% (Figure 2.13). The mode of addition in th epreparat ion of the ester enolate catalyst 2.72 and the removal of THF and i-PrOH afterthis preparation is essential for effective catalysis (Eq. 2.20). Remarkably, with MVKand m alonate 2.47 , affording a product with a stereogenic center on the 4-position, anenantioselectivity of 62% was achieved.

R

R''

O

OR' La

OO

O

( )n

[OO* O

R''

OR

OR'

La

( )n

R

O O

R'R''

O

R''

OR

OR'

( )n

2.1 or 2.2

2.47, 2.53, 2.56 or 2.57

2.72

Chapter 2

24

Figure 2.13 Proposed mechanism for the asymmetric Michael reaction catalysed bya chiral lanthanum complex.

In subs equent studies the same yields and enantioselectivities were reported by a nalkali metal containing trimeric complex, derived of La(O- i-Pr) , (R)-BINOL ( 33

equivalents) and NaO- t-Bu (3 equivalents). With this catalyst the reaction of chalcone54

(2 .5) with malonate 2.53 proceed smoothly to give the Michael product in 77% e.e .(93% yield), whereas with catalyst 2.72 only 7% e.e. was found.Very recently this heterobimetallic catalysis concept was extended with th edevelopment of an amphoteric asymmetric complex from aluminium, lithium and ( R)-BINOL. Efficient complex preparation from LiAlH and two equivalents of ( R)-BINOL55

4

in THF afforded heterobimetallic catalyst 2.73 , which gave excellent yields an denantioselectivities (91-99%) in the Michael addition of several malonates to cyclicenones 2.2 or 2.3 . Al NMR studies revealed that the carbonyl group of the enone is27

coordinated to the aluminum. This mechanistic feature gave the opportunity to trap theAl-enol ate with an electrophile. The reaction of cyclopentenone ( 2.1), diethy lmethylmalonate and 3-phenylpropanal in the presence of 10 mol% of 2.73 gave thethree-component coupling product as a single diastereomer in 91% e.e. (64% yield)(Eq. 2.21). This is the first example of a catalytic asymmetric tandem Michael-aldol55

reaction.

(2.21)

2.73

OO

Li

AlOO

91% e.e.

O

Ph

OHH

CO2EtCO2Et

H2.73 (cat.)

THF, RTPh H

O+EtO OEt

O O+ 2.1


25

Chiral amines and crown ethers as catalysts The use of chiral amines as catalyst in the Michael addition reaction was reported byLångst röm and Bergson in 1973. The addition of methyl-1-oxo-2-indanecarboxylate56

(2.45 ) to acrolein ( 2.59 ) using optical active 2-(hydroxymethyl)quinuclidine ( 2.74 ,Figure 2.14) provided optical active Michael product (see Eq. 2.18).Wynberg and co-workers studied the model reaction of the same Michael donor withMVK as Michael acceptor and quinine ( 2.75 ) as chiral base (1 mol%). The Michael57

product is produced in almost quantitative yield with e.e.'s up to 76%, depending onsolvent and temperature (Eq. 2.18). Several variations of chiral base, Michael donorand ac ceptor, and reaction conditions were examined in detail but enantioselectivitiesexceeding 76% were not reached in these studies. 58 ,59

Several a ttempts were reported to facilitate removal of the chiral catalyst from th ereac tion mixture by attaching it to a polymer. With alkaloids 2.75 , 2.77 , and 2.78anchored to cross-linked polystyrene or co-polymerized with acrylonitrile, the60 61

Michael addit ions given in Eq. 2.18 proceed with low and moderate e.e.'s, respectively.Insertion of spacer groups between the alkaloid and the polymer backbone improvesthe enantioselectivity to 65%. This is almost the same value as was found in th e62

reaction with non-polymer bound alkaloid. When the model reaction (Eq. 2.18) wasperformed under high pressure lower e.e.'s were found. 63

2.74

2.75 R = - ; X = -2.76 R = CH 2Ph ; X = F

2.79 R = C12H25 ; X = F2.80 R = C12H25 ; X = Br2.81 R = CH2Ph ; X = Br2.82 R = CH2Ph ; X = Cl

OO

O

OO

O

2.83

NOH

S

N

OMe

N

R

OH+X

_+

R

N

N

R'

HO

R X_

2.77 R' = OMe ; R = - ; X = -2.78 R' = H ; R = CH 2Php-CF3

X = Br

HO N

Ph

R+

X_

2.86 R = Me2.87 R =

MeO

CH2O

O

O

O

OO

O O

OO

O

OO

R

RO

O

O

4

2.84 2.85

Chapter 2

26

Figure 2.14 Chiral amines and crown ethers used in Michael additions.

The model reaction was also performed under phase transfer conditions. Wit hquartenary ammonium halides, derived from methionine, the reaction is sluggish andhard ly enantioselective. Significant improvements were achieved with [p -64

(trifluoromethyl)benzyl]-chinchoniniumbromide ( 2.78 ) as phase transfer catalyst and2-propylinda none as Michael donor, and under solid-liquid phase transfer conditions65

in presence of quaternary ammonium salts, 2.80 -2.82 , derived from N-methylephedrine(a typical example is given in Eq. 2.22). 66

+ (2.22)PhPh

O

N CO2Et

O

H

CO2Et

PhPh

NEtO2C CO2Et

OH O

48-57% yield60-68% e.e.

2.5

2.80-82 (cat.)KOH (cat.)

Ph CO2Me

R

+

R = H, Me

OMe

OOMe

CO2Me

RPh

O

(2.23)2.84 (cat.)base (cat.)


27

High er enantioselectivities have been reached using chiral catalysts prepared b ycomplexation of a ba se to a chiral crown ether. Cram and Sogah found that with 4 mol%of a bis- ß-naphthol derived optically active crown ether 2.83 (Figure 2.14) an dpotassium t-butoxide as the base, the Michael product (Eq. 2.18) was isolated in 48%yield with an e.e. of 99%. 67

Crown ether 2.84 was used similarly in the reaction of methyl acrylate and methyl 2-phenylpropionate or methyl phenylacetate (Eq. 2.23). The highest e.e.'s in the latter65

reactions were achieved with potassium amide as base (83% and 65% e.e. ,respectively). In both cases the R crown ether gave the S product.

In the presence of KO t-Bu, crown ethers 2.83 and 2.84 , were also used as chiral catalystin the an ionic (Michael type) polymerization of methacrylate esters to give highl yisotactic helical polymers. 68

Following thes e fascinating reports, several groups have been investigated other chiralcrown ethers as catalysts in the reaction given in Eq. 2.23. Enantioselectivities did not69

rise above 81%. A remarkable high enantioselectivity of 79% was achieved with simpleC -symmetric chiral crown ether 2.86 derived from 2 S,3S-butanediol. With chira l2

69d

cr own ether 2.85 , Yamamoto and co-workers investigated the Michael addition o fmethyl phenylthioacetate to cyclopentenone to give the Michael product in 60% yieldwith an e.e. of 41% (Eq. 2.24). With crown ether 2.87 Koga and co-workers were able69c

to enhance the enantioselectivity to 68% in this reaction. 70

AIBNBu3SnH2.85 or

2.87 (cat.)KOt-Bu(cat.)

2.1 PhS CO2Me+

SPh

CO2Me

O

CO2Me

O

(2.24)

41-68% e.e

2.79-2.81or 2.76(cat.)

R = Ph, Me

(2.25)R

O

Ph

O2N

CH3NO2+R

O

Ph

NH

R

2.88 R = CH2NHPh2.89 R = CONH22.90 R = CH2OH

Chapter 2

28

2.7 Nitroalkane additions

In recent years, Michael addition reactions of nitroalkanes to activated alkenes haveattracted considerable attention in part due to the availability of various syntheti cmetho ds for the conversion of the nitro group to other functional groups. Only a few71

enant ioselective nitromethane additions to enones catalysed by alkaloids an dder ivatives have been reported. With quaternary salts derived of quinine or N-methylephedrine (2.76, 2.79 -2.81 , Figure 2.14) as chiral phase transfer catalysts andexcess of inorgan ic salts (KF, NaOH or KO t-Bu) enantiomeric excesses up to 26% werereached in the add ition of nitromethane to chalcone (Eq. 2.25). With the free alkaloids72

as chiral bases no reaction takes place in aprotic solvents. In methanol addition takesplace although without enantioselectivity.

Under high pressure (900 MPa) both quinine ( 2.75 ) and quinidine ( 2.77 ) (10 mol%)catalyse the nitromethane addition in aprotic solvents liketolue ne, with high conversion and e.e.'s up to 60%. 63, 73

These results shows that high pressure is actuall yadvanta geous in performing sluggish asymmetric reactionscomposed of rather inert reactants and/or poor catalysts.Bo tteghi and co-workers reported the first example of atransition-metal catalysed enantioselective nitroalkan eaddition to enones (Eq. 2.25). The catalyst was prepared74

2.91 ( = L* )

2.5MeS

SSnL*(OTf)2.91 (cat.)Sn(OTf)2 (cat.)

CH2Cl2, -78°C

NNH

SSiMe3

MeS (2.26)+ MeS

S

PhPh

O


29

in situ from Ni(acac) and proline-derived ligands 2.88 -2.90 . Using a large excess of2

nitr omethane, enantioselectivities up to 17% where reached although long reactio ntimes are req uired. With benzylideneacetone slightly higher e.e.'s (24%) but low yieldswere fo und. A decrease of Michael donor-to-acceptor ratio appears to increase th easymmetric induction. With equimolar amounts of donor and acceptor the chemica lyield of the Michael product is rather low, but an enantioselectivity of 61% is found(ligand 2.89). The observed increase in e.e. might well be a solvent effect. An increase48

in solvent polarity (nitromethane vs benzene) produces a decrease in stereoselectivityin the same reaction catalysed by alkaloid bases under phase transfer conditions. 58b

Asymmetric catalysis is confirmed to be a property of metal complexes in this case asthe ligands alone do not catalyse the reaction.Yamaguchi and co-workers noted a reaction of 2-nitropropane and 2-cyclohexenone(2.2) or (E)-3-penten-2-one ( 2.61 ) in the presence of 5 mol% of rubidium salt 2.70 . 51

The products were obtained with yields of 61% and 48% and e.e.'s of 58% and 69%,respectively .

2.8 Miscellaneous

Two additional successful approaches to catalytic asymmetric Michael addition needto be mentione d, which use chiral Lewis acids as catalysts. Mukaiyama and co-workersused a c hiral tin complex derived in situ from tin(II) triflate and chiral diamine 2.91 inthe Micha el addition of trimethylsilyl enethiolate to enones (Eq. 2.26). When th e75

trimethylsilyl enet hiolate was added slowly to the reaction mixture, in order to suppressthe co mpeting uncatalysed addition, enantioselectivities up to 70% were reached. It ispropo sed that metal exchange of tin and silicium initially takes place to generate achiral tin(II) enethiolate and Me SiOTf. Activation of the enone by Me SiOTf induces3 3

the Micha el addition of the chiral tin(II) enethiolate along with the regeneration of thetin(II) triflate-diamine complex.

2.92

1) 2.92 (cat.)Cl2Ti(Oi-Pr)2 (cat.)MS 4Å, ether, 0°C

2) H3O+

O

O

PhOH

OH

PhPh

PhPh

R = Ph, t-Bu 36-68% yield53-55% e.e

(2.27)N OMeO

O O

O

O

R+ N O

MeOO O

ON

O

R

Chapter 2

30

Catalytic asy mmetric Michael additions of morpholine derived enamines to methyl( E)-4-o xo-4-(2-oxo-1,3-oxazolidin-3-yl)-2-butenoate with modest yields but promisin ge.e.'s were found by Narasaka and co-workers (Eq. 2.27). The chiral catalyst i s76

pr epared in situ from Cl Ti(O i-Pr) and (2 R,3R)-1,4-diol, 2.92 , derived from tartari c2 2

acid . The course of the titanium-catalysed addition reaction of enamines wit hunsaturated acid deri vatives was strongly depending on the enamine structure. Contraryto the Mich ael reaction of the enamines given above, the reactions of 2,2-disubstitutedenamines with the same Michael acceptor were found to afford optically activ ecyclobutanes. The in situ prepared chiral catalyst was also used successfully in Diels

Alder and [2+2] cycloadditions, ene reactions, and hydrocyanations.

2.9 Conclusions

The current stage of enantioselective synthesis of ß-substituted carbonyl compoundsusing chiral catalysts has been reviewed in this chapter. Remarkable progress has beenmade in the last few years on the enantioselective synthesis of ß-substituted carbonylcompounds by conjugate addition catalysed by chiral metal complexes. Except for anearly re port by Brunner on cobalt-catalysed Michael additions, the first successfu lenant ioselective conjugate addition reactions catalysed by metal complexes appearedin 1988. A number o f examples are currently known of both Michael type additions and1,4-addi tions of organometallic reagents catalysed by chiral metal complexes wit henantioselectivities exceeding 80%. The large variety of chiral metal complexes andligan ds that have shown modest enantioselectivities are the stepping stones for th edevelopment of highly selective catalysts in the near future.


31

1. a) Perlmutter, P. Conjugate Addition Reactions in Organic Synthesis; Tetrahedron Organic ChemistrySeries, No. 9; Pergamon: Oxford, 1992. b) Winterfeldt, E. Kontakte (Darmstadt) 1987 , 20; 1987 , 37.

2. a) Schmalz, H.-G. in Comprehensive Organic Synthesis; Trost, B.M.; Fleming, I., Eds.; Pergamon :Oxford, 1991 ; Vol 4, Chapter 1.5. b) Wynberg, H. in Topics in Stereochemistry; Eliel, E.L.; Wilen, S.,Eds.; Wiley-Interscience: New York, 1986 ; Vol. 16, p 87. c) Tomioka, K.; Koga, K. in AsymmetricSynthesis; Morrison, J.D., Ed.; Academic: New York, 1983 ; Vol. 2, p 201.

3. Rossiter, B.E.; Swingle, N.M. Chem. Rev. 1992 , 92, 771.

4. House, H.O. Modern Synthetic Reactions 2 ed.; W.A Benjamin, Inc: Menlo Park, California, 1972 ;nd

Chapter 9.

5. Feringa, B.L.; de Lange, B.; Jansen, J.F.G.A.; de Jong, J.C.; Lubben, M.; Faber, W.; Schudde, E.P. PureAppl. Chem. 1992 , 64, 1865.

6. a) Jansen, J.F.G.A.; Feringa, B.L. in Houben-Weyl, Stereoselective Synthesis of Organic Compounds;Hoffm ann, R.W.; Mulzer, J.; Schaumann, E. Eds.; Georg Thieme Verlag: Stuttgart, 1995 , Chapte r1.5.2.3. b) Nogradi, M. Stereoselective Synthesis; VCH Publ.: Weinheim, 1987 .

7. See for insta nce: a) Davies, S.G.; Walker, J.C. J. Chem. Soc., Chem. Commun. 1985 , 209. b) Liebeskind,L.S.; Welker, M.E. Tetrahedron Lett. 1985 , 26, 3079.

8. Corey, E.J.; Naef, R.; Hannon, F. J. Am. Chem. Soc. 1986 , 108, 7114.

9. Aoki, Y.; Kuwajima, I. Tetrahedron Lett. 1990 , 31, 7457 and references therein.

10. Corey, E.J.; Boaz, N.W. Tetrahedron Lett. 1985 , 26, 6015, 6019.

11. Mukaiyama, T.; Tamura, M.; Kobayashi, S. Chem. Lett. 1986 , 1017 and references therein.

12. Heathcock, C.H.; Norman, M.H.; Uehling, D.E. J. Am. Chem. Soc. 1985 , 107, 2797.

13. Van Draanen, N.A.; Arseniyadis, S.; Crimmins, M.T.; Heathcock, C.H. J. Org. Chem. 1991 , 56, 2499.

14. a) Corey, E.J.; Peterson, R.T. Tetrahedron Lett. 1985 , 26, 5025. b) Corey, E.J.; Magriotis, P.A. J. Am.Chem. Soc. 1987 , 109, 287. c) Yamaguchi, M.; Hasebe, K.; Tanaka, S. Minami, T. Tetrahedron Lett.1986 , 27, 959. d) Enders, D.; Müller, S.; Demir, A.S. Tetrahedron Lett. 1988 , 29, 643, and referencestherein.

15. a) Villacor ta, G.M.; Rao, C.P.; Lippard, S.J. J. Am. Chem. Soc. 1988 , 110, 3175. b) Ahn, K.-H.; Klassen,R.B.; Lippard, S.J. Organometallics 1990 , 9, 3178.

A wealth of information has already been gathered on the factors effecting catalyticactivit y and selectivity. A picture emerges of conjugate addition reactions being oftenextreme ly delicate and complex processes, in particular due to the appearance o fvarious catalytically active complexes (in equilibrium) during the reaction and th esensitiv ity to the conditions of the reaction. The scope of organometallic reagents ,Michael donors, a nd enones in these enantioselective processes has been limited so far.With a few exceptions model reactions have been studied only. It is evident that thedevelop ment of highly selective catalysts for conjugate addition with a broad scope isa major challenge in current asymmetric synthesis.

2.10 References

Chapter 2

32

16. For a discussion on the effect of silyl reagents on the conjugate addition of organocuprates, see: Bertz,S.H.; Miao, G.; Rossiter, B.E.; Snyder, J.P. J. Am. Chem. Soc. 1995 , 117, 11023 and references therein.

17. a) Lambert, F.; Knotter, D.M.; Janssen, M.D.; van Klaveren, M.; Boersma, J.; van Koten, G .Tetrahedron: Asymmetry 1991 , 2, 1097. b) van Klaveren, M.; Lambert, F.; Eijkelkamp, D.J.F.M.; Grove,D.M.; van Koten, G. Tetrahedron Lett. 1994 , 35, 6135. c) van Koten, G. Pure Appl. Chem. 1994 , 66,1455.

18. a) Zhou, Q.-L.; Pf altz, A. Tetrahedron Lett. 1993 , 34, 7725. b) Zhou, Q.-L.; Pfaltz, A. Tetrahedron 1994 ,50, 4467.

19. Spescha, M.; Rihs, G. Helv. Chim. Acta 1993 , 76, 1219.

20. Kanai, M.; Tomioka, K. Tetrahedron Lett. 1995 , 36, 4275.

21. Isobe, M.; Kondo, S.; Nagasawa, N.; Goto, T. Chem. Lett. 1977 , 679.

22. a) Watson, R.A .; Kjonaas, R.A. Tetrahedron Lett. 1986 , 27, 1437. b) Kjonaas, R.A.; Hoffer, R.K. J. Org.Chem. 1988 , 53, 4133.

23. Jansen, J.F.G.A.; Feringa, B.L. Tetrahedron Lett. 1988 , 29, 3593.

24. Jansen, J.F.G.A.; Feringa, B.L. J. Chem. Soc., Chem. Commun. 1989 , 741.

25. Jansen, J.F.G.A.; Feringa, B.L. J. Org. Chem. 1990 , 55, 4168.

26. Tanaka, K.; Matsui, J.; Suzuki, H. J. Chem. Soc., Perkin Trans. I, 1993 , 153.

27. Tomioka, K.; Shindo, M.; Koga, K. Tetrahedron Lett. 1993 , 34, 681.

28. For a d iscussion about ligand-accelerated asymmetric catalysis, see: Berrisford, D.,J.; Bolm, C. ;Sharpless, K.B. Angew. Chem., Int. Ed. Engl. 1995 , 34, 1059.

29. For reviews see: a) Noyori, R.; Kitamura, M. Angew. Chem., Int. Ed. Engl. 1991 , 30, 49. b) Soai, K.;Niwa, S. Chem. Rev. 1992 , 92, 833.

30. a) Boersma, J. In Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F.G.A.; Abel, E.W.,Eds.; Pergamon: Oxford, 1982 , Vol. 2, Chapter 16. b) Knochel, P.; Singer, R.D. Chem. Rev. 1993 , 93,2117.

31. Greene, A.E.; Lansard, J.-P.; Luche, J.-L.; Petrier, C. J. Org. Chem. 1984 , 49, 931.

32. Soai, K.; Hayasaka, T.; Ugajin, S.; Yokoyama, S. Chem. Lett. 1988 , 1571.

33. Soai, K.; Yokoyama, S.; Hayasaka, T.; Ebihara. K. J. Org. Chem. 1988 , 53, 4148.

34. Soai, K.; Hayasaka, T.; Ugajin, S. J. Chem. Soc., Chem. Commun. 1989 , 516.

35. Bolm, C.; Ewald, M. Tetrahedron Lett. 1990 , 31, 5011.

36. Bolm, C.; Ewald, M.; Felder, M. Chem. Ber. 1992 , 125, 1205.


38. Bolm, C.; Felder, M.; Müller, J. Synlett 1992 , 439.

39. Corma, A.; Iglesias, M.; Martín, V.; Rubio, J.; Sánchez, F. Tetrahedron: Asymmetry 1992 , 3, 845. Forthe synthesis of 2.39 and 2.40 , see: Corma, A.; Iglesias, M.; del Pino, C.; Sánchez, F. J. Organomet.Chem. 1992 , 431, 233.

40. Uemura, M.; Miyake, R.; Nakayama, K.; Hayashi, Y. Tetrahedron: Asymmetry 1992 , 3, 713.


33

41. Asami, M.; Usui, K.; Higuchi, S.; Inoue, S. Chem. Lett. 1994 , 297.

42. Fujisawa, T.; Itoh, S.; Shimizu, M. Chem. Lett. 1994 , 297.

43. Soai, K.; Okudo, M.; Okamoto, M. Tetrahedron Lett. 1991 , 32, 95.

44. Alexakis, A.; Frutos, J.; Mangeney, P. Tetrahedron: Asymmetry 1993 , 4, 2427.

45. Brunner, H.; Hammer, B. Angew. Chem., Int. Ed. Engl. 1984 , 23, 312.

46. Brunner, H.; Kraus, J. J. Mol. Cat. 1989 , 49, 133.

47. a) Desimoni, G.; Quadrelli, P.; Righetti, P.P. Tetrahedron 1990 , 46, 2927. b) Desimoni, G.; Dusi, G.;Faita, G.; Quadrelli, P.; Righetti, P. Tetrahedron 1995 , 51, 4131.

48. Botteghi, C.; Paganelli. S.; Schionata, A.; Boga, C.; Fava, A. J. Mol. Cat. 1991 , 66, 7.

49. Sawamura, M.; Hamashima, H.; Ito, Y. J. Am. Chem. Soc. 1992 , 114, 8295.

50. Sawamura, M.; Kuwano, R.; Ito, Y. Angew. Chem., Int. Ed. Engl. 1994 , 33, 111.

51. a) Yamaguchi, M.; Shiraishi, T.; Hirama, M. Angew. Chem., Int. Ed. Engl. 1993 , 32, 1176. b )Yamaguchi, M.; Shiraishi, T.; Hirama, M. J. Org. Chem. 1996 , 61, 3520.

52. Kawara, A.; Taguchi, T. Tetrahedron Lett. 1994 , 35, 8805.

53. Sasai, H.; Arai, T.; Shibasaki, M. J. Am. Chem. Soc. 1994 , 116, 1571.

54. Sasai, H.; Arai, T.; Satow, Y.; Houk, K.N.; Shibasaki, M. J. Am. Chem. Soc. 1995 , 117, 6194.

55. Arai, T.; Sasai, H.; Aoe, K.; Okamura, K.; Date, T.; Shibasaki, M. Angew. Chem., Int. Ed. Engl. 1996 ,35, 104.

56. Långström, B.; Bergson, G. Acta Chem. Scand. 1973 , 27, 3118.

57. Hermann, K.; Wynberg, H. J. Org. Chem. 1979 , 44, 2238.

58. a) Wynberg, H.; Helder, R. Tetrahedron Lett. 1975 , 4057. b) Wynberg, H.; Greijdanus, B. J. Chem. Soc.,Chem. Commun. 1978, 427. c) Kobayashi, N.; Iwai, K. J. Pol. Sci., Pol. Lett. Ed. 1982 , 20, 85. d) Heisler,T.; Janowski, K.; Prager, R.H.; Thompson, M.J. Aust. J. Chem. 1989 , 42, 37.

59. For related enantioselective Michael addition of "-i-propyl-3,4-dimethoxybenzyl cyanide to severaleno nes with e.e. < 11%, see: Brunner, H.; Zintl, H. Monatsh. Chem. 1991 , 122, 841. For conjugat ecyanide addition to enones with e.e. < 45%, see: Dehmlow, E.V.; Sauerbier, C. Liebigs Ann. Chem.1989 , 181.

60. a) Hermann, K.; Wy nberg, H. Helv. Chim. Acta 1977 , 60, 2208. b) Hodge, P.; Khoshdel, E.; Waterhouse,J. J. Chem. Soc., Perkin Trans. I 1983 , 2205.

61. a) Kobayashi, N.; Iwai, K. J. Am. Chem. Soc. 1978 , 100, 7071. b) Kobayashi, N.; Iwai, K.; J. Pol. Sci.,Pol. Chem. Ed. 1980 , 18, 923.

62. Inagaki, M.; Hiratake, J.; Yamamoto, Y.; Oda, J. Bull. Chem. Soc. Jpn. 1987 , 60, 4121.

63. Sera, A.; Takagi, K.; Katayama, H.; Yamada, H.; Matsumoto, K. J. Org. Chem. 1988 , 53, 1157.

64. Banfi, S.; Cinquini, M.; Colonna, S. Bull. Chem. Soc. Jpn. 1981 , 54, 1841.

65. Conn, R.S.E.; Lovell, A.V.; Karady, S.; Weinstock, L.M. J. Org. Chem. 1986 , 51, 4710.

66. Loupy, A.; Sansoulet, J.; Zaparucha, A.; Merienne, C. Tetrahedron Lett. 1989 , 30, 333.

67. Cram, D.J.; Sogah, D.Y. J. Chem. Soc., Chem. Commun. 1981 , 625.

Chapter 2

34

68. Cram, D.J.; Sogah, D.Y. J. Am. Chem. Soc. 1985 , 107, 8301.

69. a) Alonso-López, M.; Martín-Lomas, M.; Penadés, S. Tetrahedron Lett. 1986 , 27, 3551. b) Alonso-López, M.; Jiminez-Barbero, J.; Martín-Lomas, M.; Penadés, S. Tetrahedron 1988 , 44, 1535. c) Takasu,M.; Wakabayashi, H.; Furuta, K.; Yamamoto, H. Tetrahedron Lett. 1988 , 29, 6943. d) Aoki, S.; Sasaki,S.; Koga, K. Tetrahedron Lett. 1989 , 30, 7229. e) Maarschalkerwaart, D.A.H.; Willard, N.P.; Pandit,U.K. Tetrahedron 1992 , 48, 8825. f) Crosby, J.; Stoddart, J.F.; Sun, X.; Venner, M.R.W. Synthesis, 1993 ,141.

70. Aoki, S.; Sasaki, S.; Koga, K. Heterocycles, 1992 , 33, 493.

71. a) Seebach, D.; Colvin, E.W.; Lehr, F.; Weller, T. Chimia 1979 , 33, 1. b) Ono, N.; Kaji, A. Synthesis1986, 693. c) Seebach, D.; Missbach, M.; Calderari, G.; Eberle, M. J. Am. Chem. Soc. 1990 , 112, 7625.

72. a) Colonna, S.; Hiemstra, H.; Wynberg, H. J. Chem. Soc., Chem. Commun. 1978 , 238. b) Annunziata,R.; Cinquini, M.; Colonna, S. Chem. Ind. 1980 , 238. c) Colonna, S.; Re, A.; Wynberg, H. J. Chem. Soc.,Perkin Trans. I 1981 , 547.

73. Matsumoto, K.; Uchida, T. Chem. Lett. 1981 , 1673.

74. Schionata, A.; Paganelli, S.; Botteghi, C.; Chelucci, G. J. Mol. Cat. 1989 , 50, 11.

75. a) Yura, T.; Iwasawa, N.; Narasaka, K.; Mukaiyama, T. Chem. Lett. 1988 , 1025. b) Iwasawa, N.; Yura,T.; Mukaiyama, T. Tetrahedron 1989 , 45, 1197.

76. Hayashi, Y.; Otaka, K.; Saito, N.; Narasaka, K. Bull. Chem. Soc. Jpn. 1991 , 64, 2122.

PhPh

O

+ Et2Zn

3.1a 3.2a

N

OHN(alkyl)2

Ph Me

HOtBu

OH

Ar N

Ph Ph

O

*chiral ligand (cat.)CH3CN

Ni(acac)2 (cat.)

3.3 3.4 3.5a

1

Chapter 3

Conjugate Addition of Diethylzinc to Chalcones Catalysed by Chiral Nickel(II) Amino Alcohol Complexes

3.1 Introduction

A major synthetic challenge is the development of methodology for catalyti cenan tioselective carbon-carbon bond formation. Among these, the catalyti cena ntioselective addition of dialkylzinc reagents to aldehydes has been studie dex tensively. Successful reports on chiral catalysts for conjugate addition reaction s1

are less frequent, as is discussed in Chapter 2. In this Chapter the nickel catalyse denanti oselective conjugate addition of diethylzinc to chalcone will be described i ndetail . A brief survey of the results reported so far is required. Since Greene an dLu che and co-workers have shown that organozinc reagents have large potential i nthe nickel catalysed conjugate addition reaction, a number of groups have focussed2

their attention on enantioselective versions of this carbon-carbon bond formatio n(see Section 2.5). Soai and co-workers developed an enantioselective modificatio nof the nickel catalysed alkyl transfer from diethylzinc to chalcone ( 3.1a ) with N,N-dialkyln orephedrine (3.3) as chiral ligand (Scheme 3.1). Similar results wer e3

obtai ned by Bolm and co-workers using C -symmetric bipyridines or substitute d2

chiral pyridines 3.4 as ligands. Jansen in our group has found that comparabl e4

enantioselectivities are reached with cis-exo-N,N-dimethyl-3-aminoisoborneo l(3.5a ) as chiral ligand. 5

Scheme 3.1 Asymmetric conjugate addition of diethylzinc to chalcone

Chapter 3

2

(3.1a).

While the se efforts represent a significant advance in this field, there is yet n ogeneral solution to the problem of achieving efficient catalytic enantioselectiv econ jugate addition to a wide range of enones. Furthermore, at the start of thi sin vestigation, detailed knowledge about the factors governing catalytic activity ,regio-, an d enantioselectivity were not known. In this Chapter details of ou rinvestigation on the nickel / amino alcohol catalysed conjugate addition o fdialkylz inc reagents to enones will be described. In order to assess the differen t6

variab les, the 1,4-addition of diethylzinc to chalcone has been examine dextensively.

3.2 Preliminary experiments

In o rder to find a suitable ligand, which could be modified systematically t ooptimise the enantioselectivity, 3.5a and nine other chiral ligands (Figure 3.1) wereexamined in the nickel catalysed addition of diethylzinc to 3.1a (the model reaction,Scheme 3.1). Compounds 3.6 , 3.7 , 3.8 , and 3.14 have been successfully applied a schiral catalysts in the 1,2-addition of dialkylzinc to benzaldehyde. However, th e1

ability of these compounds to induce asymmetry in a conjugate fashion is unknown.Compounds 3.9 , 3.10 , and 3.11 have been synthesised from L-valine by Knol and7

compounds 3.12 and 3.13 have been prepared by Stock in this laboratory and ar e8

easily available for examination.

OHOH

N

OH

N

N OH

PhPh

CF3SO2N OHH

RR

R = 2,4,6-tri-methylphenyl

RSO2N OHH

2-naphthyl2-naphthyl

3.10 R = CF33.11 R = p-tolyl

3.6

3.7 3.8 3.9

N

HOH

N

OCH3

3.12

OO

SS

3.13

OCH3

OCH3

SS

SS

3.14


3

Figure 3.1 Structures of the chiral ligands 3.6-3.14.

The catalytic conjugate addition reactions were run on a 1 mmol-scale. In general, aso lution of 7 mol% Ni(acac) and 16 mol% chiral ligand in 2 ml of acetonitrile wa s2

hea ted at reflux for 1 h, where upon substrate was added at room temperatur efollowed by 1.5 ml of a solution of diethylzinc in hexane (1 M) at -30 EC. Th eprod uct 1,3-diphenylpentan-1-one ( 3.2a ) was isolated by standard techniques an dthe enantiomeric excess was determined by chiral HPLC analysis (experimenta lsection). The results of these reactions are shown in Table 3.1. Excellent chemica lyields of 3.2a were obtained in most cases. When thiacrownether ligands 3.12 or3.13 wer e employed yields were slightly lower and reduced byproducts (1,3 -diphenylpropan-1-one and 1,3-diphenyl-2-propenol) have been detected (< 5%) .Apparently, the alkyl addition is somewhat hampered by the huge crown ethers andthe undesir ed reduction, possible if hydrogen atoms $ to the metal are present ,beco mes competing. Significant enantiomeric excesses (e.e.) were only obtaine d2b

for (-)-cis-exo-N,N-dimethyl-3-aminoisoborneol ((-)-DAIB) ( 3.5a ) and (+) -9,10 ,11

di phenyl-(1-methylpyrrolidin-2-yl)methanol ((+)-DPMPM) ( 3.8). Although nicke l12

complexe s with ligands containing mixed donor atoms (in general N, O and S) ar einnu merable, it should be emphasised that both these enantioselective ligands ar e13

Chapter 3

4

tertiary ß-amino alcohols. Bis-ß-naphthol ( 3.6), (2S,2'S)-2-(hydroxymethyl)-1-[(1 -met hylpyrrolidin-2-yl)methyl]pyrrolidine (3.7), ß-hydroxy-sulfonamides 3.9-3.11 ,mi xed crown ethers 3.12 and 3.13 and quinine ( 3.14 ) gave low enantioselectivities ,indica ting that very specific ligand properties are required for enantioselectiv enickel catalysed conjugate addition of diethylzinc to chalcone.

Table 3.1 Nickel catalysed enantioselective conjugate addition of diethylzinc tochalcone (3.1a) (Scheme 3.1).a

entry ligand yield (%) e.e. (%) abs. conf.b c d

1 3.5a 94 59 R 2 3.6 87 0 - 3 3.7 92 8 R 4 3.8 94 25 R 5 3.9 97 6 R 6 3.10 96 2 R 7 3.11 98 2 R 8 3.12 63 2 R 9 3.13 76 3 R 10 3.14 89 5 S

a. Reactions at -30 EC in 2 ml of acetonitrile and 1.5 ml of hexane using an in situ prepared catalyst from 7mol% of Ni(acac) and 16 mol% of chiral ligand (see text and experimental section). Reaction time 16 h. b.2

Isolat ed yield. Conversion to the 1,4-product > 95% (based on GC analysis). c. Determined by HPL Canalysis: Daicel, Chiralcel OD; 0.25% iPrOH in hexane, flow rate 1.0 ml/min, UV detector (254 nm). d .Comparison of retention times of 3.2a with known data gave the absolute configuration. 4c

Since (- )-DAIB gave the highest e.e. in the nickel catalysed conjugate addition w ehave used this ligand for further study. First the effect of additional achiral amin eligands, which showed to be essential to obtain high e.e.'s in a related system, was3c

exam ined. With piperidine or 2,2'-bipyridine (1 equivalent with respect to nickel )the e.e. was enhanced to 70% and 85%, respectively, in the addition of diethylzinc tochalcone (Scheme 3.1). It should be emphasised that after careful examination o f5

the ch iral ligand employed it appeared that slightly polluted (-)-DAIB was use d(vide infra). Before further investigation of the catalytic reaction we focussed ou rattenti on on the synthesis of (-)-DAIB and several other (+)-camphor-derive doptically pure tertiary amino alcohol ligands.

3.15O

NOH

RN

R

OH NOH

RRN

HO RR

3.5a R = CH33.5b R = Hd

3.16a R = CH33.16b R = Hd

3.17a R = CH33.17b R = Hd

a b c


5

3.3 Synthesis of chiral amino alcohols derived from (+)-camphor

(+)-Camphor, readily available from natural sources or produced synthetically frompinene , has found extensive use as chiral building block in natural produc tsynthe sis. Derivatives of camphor have been used frequently as chiral auxiliarie s14

in asy mmetric synthesis. The most striking example is probably (-)- cis-exo-N,N-15

di methyl-3-aminoisoborneol ((-)-DAIB) ( 3.5a ), introduced by Noyori, as excellen tchiral catalyst in the addition of dialkylzinc reagents to several aldehydes. 1a,11

Scheme 3.2 Synthesis of three stereoisomers of $-amino alcohols derived from (+)-camphor. (a) LiAlH , diethyl ether (b) Zn, NaOH, H O; LiAlH , diethyl4 2 4

ether (c) Na, ethanol (d) MeI (2.5 equivalents), NaOH, diethyl ether /H O.2

In gen eral, (-)-DAIB is synthesised according to a procedure published b yCh ittenden and Cooper (Scheme 3.2). Direct reduction of 3-hydroxyiminobornan -10

2-one (3.15 ), easily prepared from (+)-camphor and isoamyl nitrite, with LiAlH164

gives stereoselective cis-exo-3-aminoisoborneol ( 3.5b ). Double methylation of 3.5bby slight excess of methyl iodide in diethyl ether proceeded readily to the tertiar yamino alcohol (-)-DAIB ( 3.5a ), while quaternisation was found to be negligible. 17

HONH2

ONH

O

HONH2

3.16b 3.16b3.18

a b

(crude) (pure)

Chapter 3

6

Ho wever, (-)-DAIB was contaminated with small amounts of (+)- cis-endo-N,N-dime thyl-3-aminoborneol ((+)-DAB, 3.16a ). Pure (-)-DAIB can either be obtaine dby crystallisation with equimolar amount of (2 R,3R)-(+)-tartaric acid or b y18

successive distillation, column chromatography and another distillation.Pr imary amino alcohols cis-endo-3-aminoborneol ( 3.16b ) and trans-3-aminoisoborneol (3.17b ) were prepared from 3.15 by known reducin gproced ures. Again both amino alcohols were contaminated with othe r10, 19

stere oisomers. Purification was possible after conversion with methyl iodide to th etertiary amino alcohols 3.16a and 3.17a .Other possible routes to pure primary amino alcohols ( 3.5b and 3.16b ) derived from(+)-camphor are a highly stereoselective reduction of 3.15 using alkylaluminiu mdichlorides or via crystallisation of cyclic cis-carbamates (Scheme 3.3).20 21

Scheme 3.3 Purification of crude cis-endo-3-aminoborneol by derivatisation tothe cyclic carbamate, crystallisation and hydrolysis. (a) diethyl20

carbonate, K CO (b) NaOH, ethanol / H O.2 3 2

Since many asymmetric reactions require steric bulk in order to reach hig hena ntioselectivities we synthesised, according to standard N-alkylating procedures ,N-mon o- and N-dialkylated cis-exo-aminoisoborneols 3.5c-g (Scheme 3.4). N-meth yl-3-aminoisoborneol ((+)-MAIB) ( 3.5c ) was prepared from 3.5b according t oa literature procedure. Unfortunately, several attempts with ethyl iodide t o22

synt hesise N,N-diethyl-3-aminoisoborneol failed. Using for example the condition sgiven in Scheme 3.4 only the mono substituted aminoisoborneol 3.5d was obtained,probably due to steric hindrance.The alkylation of 3.5b with slight excess of 1,3-dibromopropane, 1,4 -dibromobutane or 1,5-dibromopentane and K CO as base in refluxing ethanol gave2 3

novel ligands 3.5e-g . Yields of the pure compounds were moderate (40-50%) partlyas a result of N-monoalkylation and extensive purification. In order to create a naziridine substituted isoborneol, 1,2-dibromoethane was used in this syntheti c

NH2

OH

3.5b

3.5c R = Me

3.5e n = 13.5f n = 23.5g n = 3

a

b

NHOH

R

NOH

( )n

3.5d R = Etb


7

proce dure. Unfortunately, only unidentified material or starting material has bee nisolated. Later on in this project, the synthesis of 3.5g was performed in refluxing acetonitrile(yield 80%). So, K CO in refluxing acetonitrile seems to be the optimal conditions2 3

for sterically hindered N-alkylating procedures.

Scheme 3.4 Synthesis of mono and dialkylated cis-exo-aminoisoborneols. (a) i.formamide; ii. LiAlH , THF (b) corresponding alkyl(di)halide, K CO ,4 2 3

ethanol.

For (-)-DAIB and the new ligand 3.5e the enantiomeric purity was determined by P31

NMR (in cooperation with Hulst). Derivatisation of racemic 3.5a or 3.5e with ach iral trivalent diazaphospholidine ( 3.24 ) in CDCl gave two base line separate d3

signals in the decoupled P NMR spectra (structure of 3.24 and procedure, se e31

exper imental section). With enantiomeric pure 3.5a or 3.5e only one signal coul d23

be determined.

3.4 (+)-Camphor-derived $$-amino alcohols as ligands in the nickelcatalysed addition of diethylzinc to chalcone. Optimisation of thecatalyst composition

Wh en we employed completely pure (-)-DAIB as chiral ligand in the nicke lcataly sed conjugate addition of diethylzinc to chalcone ( 3.1a , Scheme 3.1) the e.e .of 3.2a rose from 59% to 65%. However, the previously observed promoting effec tof addi tional achiral ligands on the e.e. of 3.2a had been disappeared (Table 3.2 ,

Chapter 3

8

entries 1-4 ). A slight colour change from green to brown was observed when 2,2' -bipyridine was added, but no significant change in yield or e.e was seen.Probably (+)-DAB or another contamination in combination with (-)-DAIB and 2,2'-bipyridine was responsible for the substantial improvement of the enantioselectivityobserve d in the preliminary experiments. Therefore, experiments were performe d24

in wh ich (-)-DAIB was deliberatily contaminated with small amounts of the othe rstere oisomer. However, when a combination of 15 mol% (-)-DAIB, 2 mol% (+) -DAB, 7 mol% of 2,2'-bipyridine and 7 mol% of Ni(acac) was used as chiral catalyst2

in the mo del reaction, 3.2a with only 55% e.e. was isolated (instead of 85%). Als oother ratios of (-)-DAIB and (+)-DAB were less enantioselective, indicating that thehighly enantioselective ligand mixture employed in the preliminary experiment scould not be reproduced.A second pos sible explanation for the observed enhancement of e.e of 3.2a in th epr eliminary experiments is contamination of aromatic compounds during the e.e .det ermination, furnishing an invalid value. Nowadays, the absorptions of bot henantiomers is verified by UV spectra from 200-300 nm.

Table 3.2 (+)-Camphor derived $-amino alcohols as ligands in the Ni(acac) 2

catalysed conjugate addition of diethylzinc to chalcone (Scheme 3.1). a

entry ligand (mol%) add. ligand (mol%) yield (%) e.e. (%) abs. conf.b b b

1 3.5a (16) - 81 65 R 2 3.5a (16) 2,2'-bipyridine (7) 82 64 R 3 3.5a (16) 2,2'-bipyridine (12) 86 63 R 4 3.5a (20) 2,2'-bipyridine (7) 90 69 R 5 3.5a (15), 3.16a (2) 2,2'-bipyridine (7) 74 55 R 6 3.16a (16) - 82 82 S 7 3.16a (16) 2,2'-bipyridine (7) 87 82 S 8 3.17a (16) - 82 4 R 9 3.5b (16) - nd 0 -c

10 3.5c (16) - 85 4 R11 3.5d (16) - 8 4 7 R

12 3.5e (16) - 78 33 R13 3.5f (16) - nd 35 Rc

14 3.5g (16) - 79 85 Rd


9

a. Reaction conditions, see Table 3.1 (unless stated otherwise). 7 Mol% of Ni(acac) is used. b. See Table 3.1.2

c. nd: yield is not determined, conversion to the 1,4-product > 95% (based on GC analysis). d. This reactionhas been performed two years after all other experiments described in this Chapter. Main differences are thesynthesis of 3.5g (in refluxing acetonitrile, see text) and the use of diethylzinc as a solution in toluene.

Whe n we employed pure (+)-DAB ( 3.16a ) as chiral ligand in the model reaction o fScheme 3.1, 3.2a was isolated in a good yield with the S enantiomer in 82% exces s(Table 3.2, e ntry 6). Again no change in yield or e.e. was observed when a nadditional achiral ligand was employed (entry 7). Thus, with two different tertiary $-amino alcohols ((-)-DAIB and (+)-DAB), both derived from naturally occurring (+)-camphor, the two enantiomers of 3.2a could be isolated with fairly high e.e.'s.Trans-N,N-dimethyl-3-aminoisoborneol (3.17a ) gave a nearly racemic 1,4-product ,indica ting the necessity of a cis configuration of the alcohol and the amine group s(entry 8). Furthermore, the N,N-dialkylamino group was found to be essential fo renanti oselectivity, because the primary aminoisoborneol 3.5b and the secondar yaminoisoborneols 3.5c and 3.5d furnished product 3.2a with very low e.e.'s (entrie s9-11) . Unfortunately, the novel N,N-dialkylated aminoisoborneols 3.5e and 3.5fwe re less enantioselective than 3.5a (entries 12 and 13). Probably azetidinyl- an dpyrro lidinyl-isoborneol are not sufficiently flexible to create an enantioselectiv enickel catalyst.With ligand 3.5g , on the other hand, a remarkable enantioselective catalyst wa sobtained (85% e.e.). Apparently, the more flexible piperidyl group in ligand 3.5g iscapable of blocking one side of the in situ prepared nickel complex furnishing ahigh ly enantioselective catalyst. It should be noted that this experiment wa spe rformed two years later than all other experiments described in this Chapter. Th ecatalysed conjugate addition was executed with a solution of diethylzinc in toluen ein stead of hexane. It would be of interest to synthesise other amino alcohols wit hst eric demanding groups on the amine moiety, using the alkylating procedure i nrefluxing acetonitrile ( vide supra).

Optimisation of the catalyst compositionWi th (-)-DAIB as chiral ligand several experiments were performed to optimise th ecatalytic composition. The e.e. of 3.2a showed to be dependent on the ligand-to -nickel ratio and the concentration of the enantioselective catalyst. With 8 mol% o fNi(acac) and 16 mol% of chiral ligand the e.e. was lowered to 55% (Table 3.3, entry2

1). Decre asing the amount of Ni(acac) to 3 or 1 mol% raised the enantiomeri c2

excess to 70% and 72%, respectively. When we decreased the concentration o f

Chapter 3

10

chiral ligand and Ni(acac) , keeping the ligand-to-nickel ratio constant, the e.e. o f2

3.2a dropped dramatically (entries 4 and 5). Thus, with low concentrations of th eenantioselective catalyst (NiL or higher aggregates Ni L , see equations 3.1 an d* *

2 m n

3.2), prepared in situ, probably other complexes are formed furnishing non-selectiveca talysis. Furthermore, competing catalysis by Ni(acac) (or Ni(acac)L ) is2

*

suppressed by an appropriate ligand-to-nickel ratio. In this study a ligand-to-nicke lratio of at least two is required. Probably, the asymmetric induction depends on th eequilibrium between the enantioselective nickel-ligand complexes (NiL and Ni L )* *

2 m n

and catalytically active nickel species (Ni(acac) and Ni(acac)L ) producing racemic2*

material or product with lower e.e. (Eq. 3.1 and 3.2).

Ni(acac) + 2 L W Ni(acac)L + L W NiL (3.1)2 2* * * *

m Ni(acac) + n L W Ni L (3.2)2 m n* *

Table 3.3 Effect of the catalyst composition on the e.e. of 3.2a.a

entry ligand (mol%) nickel salt (mol%) yield (%) e.e. (%) abs. conf.b b b

1 3.5a (16) Ni(acac) (8) 82 55 R2

2 3.5a (16) Ni(acac) (3) 79 70 R2

3 3.5a (16) Ni(acac) (1) 69 72 R2

4 3.5a (2) Ni(acac) (0.4) 69 31 R2

5 3.5a (0.2) Ni(acac) (0.04) 71 6 R2

6 3.5a (10) - n d 6 Sc d

7 3.5a (20) - nd 15 Sc d

8 3.5a (50) - nd 21 Sc d

9 3.5a (20) - nd 16 Se d

10 3.5a (16) NiBr (7) 80 39 R2

a. Reaction conditions, see Table 3.1 (unless stated otherwise). b. See Table 3.1. c. With 0.5 mmol o fchalcone (3.1a), reaction time 4 days in 3.5 ml of hexane at room temperature. d. nd: yield is not determined,conversion to the 1,4-product > 95% (based on GC analysis). e. With 0.5 mmol of 3.1a , reaction time 6 daysin 2 ml of acetonitrile and 1.5 ml of hexane at room temperature.

NH

O

NHR

OO

Ni

+ PF6

_

3.19a R = t-Bu3.19b R = (CH2)3Si_{Sup}


11

This assumption is confirmed by the observations found by Sánchez and co -worke rs. They used in the conjugate addition of diethylzinc to chalcone (Schem e3.1) isolated and fully characterised nickel complexes 3.19a and 3.19b , where a nacetylacetonate anion has been replaced by one equivalent of N-alkylaminocar -bonylpyr rolidine. The relatively high e.e.'s observed for 3.2a (77% and 95% ,25

respectively), with a chiral ligand-to-nickel ratio of 1, compared to ratios of mor ethan 2 in other studies, were explained by the fact that a single chiral complex i s3-6

used. Competing catalysis by Ni(acac) presumably cannot take place.2

Figure 3.2 Isolated and fully characterised chiral nickel complexes, which wereused successfully in the conjugate addition reaction of diethylzinc tochalcone.25

Thus, in our study the catalyst should be composed of 1 equivalent of Ni(acac) and2

2.2 equivalents of chiral ligand, furnishing a catalytic addition with acceptabl eenantioselectivity at relatively low ligand-to-nickel ratio.Next, the effect of the chiral ligand itself on the 1,4-addition of diethylzinc t ochalcone was examined. Without the nickel salt, but in the presence of 10, 20, or 50mol% of 3.5a 100% conversion of 3.1a is only achieved after 4 days (in hexane) or6 days (in acetonitrile / hexane) at room temperature. Compared to the reaction timeof 16 h (at -30 EC) for the nickel / chiral ligand catalysed reaction, this ligan dmediated addition is very slow. Furthermore, the S enantiomer was formed in a smallexcess (e.e. < 21%, Table 3.3, entries 6-9). The use of NiBr instead of Ni(acac) resulted in a relative slow conversion of 3.1a2 2

(90% after 16 h) and moderate e.e. of 3.2a (entry 10). The slow conversion an drelativ e low e.e. are probably consequences of solubility problems of NiBr in2

acetonitrile. See Section 5.2. for conjugate addition reactions of diethylzin ccatalysed by other metal salts.

3.23.1

e

H

Cl

d

Cl

H

c

H

OCH3

b

OCH3

H

a

H

H

3.1 / 3.2

R

R'

7 mol% Ni(acac) 2

Et2Zn, CH 3CN

16 mol% (-)-DAIB

O

R R'

*

O

R R'

Chapter 3

12

3.5 Variation of substrate, solvent, and reagent

Substrate variationWit h (-)-DAIB as chiral ligand we varied the substrate, solvent, and temperature o fre action. Substituted chalcones 3.1b-e were alkylated with good conversions to th e1,4-pro duct using the standard reaction conditions (Scheme 3.5). Methoxy an dchloro substituents at both aromatic rings are tolerated and e.e.'s in the range of 51%to 61 % were achieved (Table 3.4, entries 1-4). To ensure the correct assignment o fthe HPLC signals of 3.2 , racemic products were synthesised in an analogous mannerby achi ral Ni(acac) catalysis and analysed by the same method. When 3 -2

4c

nitrochalcone ( 3.1f ) was used no 1,4-addition occurred (entry 5).

Scheme 3.5 Enantioselective addition of diethylzinc to substitutedchalcones.

3.21

CH3CN, -30°C

Ni(acac)2 (cat.)3.5a (cat.)

Et2Zn+

3.20

(3.3)

OO


13

Table 3.4 Variation of substrate, solvent, temperature, and reagent. a

entry substrate solvent temp. ( EC) yield (%) e.e. (%) abs. conf.b b c

1 3.1b acetonitrile -30 nd 57 Rd

2 3.1c acetonitrile -30 nd 51 Rd

3 3.1d acetonitrile -30 nd 61 Rd e

4 3.1e acetonitrile -30 nd 59 Rd

5 3.1f acetonitrile -30 < 10 - - 6 3.1a butyronitrile -30 nd 69 Rd

7 3.1a butyronitrile -50 84 80 R 8 3.1a propionitrile -50 77 72 R 9 3.1a isobutyronitrile -50 79 72 R10 3.1a butyronitrile -50 82 85 Sf

11 3.1a acetonitrile -30 nd 59 Rg d

a. With 16 mol% of ligand 3.5a , reaction conditions, see Table 3.1 unless stated otherwise. b. See Table 3.1.c. Com parison of retention times of 3.2b-e and 1,3-diphenylbutan-1-one ( 3.2g) with known date gave th eabsolute configuration. d. nd: yield is not determined, conversion to the 1,4-product > 95% (based on G C4c

analysis). e. HPLC signals not base line separated. f. With 16 mol% of 3.16a . g. With 1.5 ml of dimethylzincin toluene (1 M). H NMR and C NMR data of 3.2g were in good agreement with the data found in th e1 13

literature. 4c

Unfortu nately, like all other known nickel-derived catalysts, this catalytic syste mshows only enantioselectivity in the conjugate addition of diethylzinc to chalcones.For example, with (-)-DAIB as ligand the nickel catalysed addition of diethylzinc to2-c yclohexen-1-one (3.20 ) gives the 1,4-product 3.21 in 80% yield, but n oenantioselectivity was observed (Eq. 3.3).

Variation of solvent and reaction temperatureStudies of the reaction conditions show that the use of a nitrile as solvent i sesse ntial for high enantioselectivity. Similar observations were made by Soai an d

Chapter 3

14

Bolm. With (-)-DAIB as chiral ligand the same degree of asymmetric induction i n3,4

the conjugate addition was found in butyronitrile instead of acetonitrile (69% an d65% e.e., respectively). A decrease of the reaction temperature from -30 EC to -50 ECdid enha nce the enantioselectivity up to 80% (Table 3.4, entries 7-9). When 1 6mol % of (+)-DAB in butyronitrile at -50 EC was employed the S enantiomer of 3.2acould be obtained with 85% e.e. (entry 10).

Reagent variationNowad ays, there are two main reasons that diorganozinc reagents, among th evarious organometallic compounds, are highly preferred as alkyl donor for catalyticenantioselective alkylation:- Due to the high covalent character of the carbon-zinc bond and a linea r

str ucture, dialkylzinc reagents react extremely sluggish with ( ",$-uns aturated) carbonyl compounds. However, effective catalysis can b eachieved with several (chiral) ligands and transmetalation to transition-metal1

compl exes. This "ligand- acceleration" is essential for successfu l3-6,25, 26 ,27

enantioselective catalysis. 28

- Organozinc reagents (RZnX and R Zn), unlike Grignard reagents an d2

orga nolithium reagents, tolerate the presence of a wide range of functiona lgroups , which opens routes to introduce alkyl units containing functiona lgroups in an enantioselective manner. 26,27

No majo r change in enantioselectivity was observed when dimethylzinc in toluen eins tead of diethylzinc in hexane was used in the conjugate addition to chalcon e(Table 3.4, entry 11). When dipentylzinc, prepared from the corresponding Grignardco mpound according to the procedure by Seebach and co-workers, was employe d27

in the nickel / (+)-DAB catalysed conjugate addition to chalcone the corresponding1,4-product 3.2h was isolated with a relatively high e.e. (70% yield, 75% e.e.). Withdioctylzinc, prepared according to the elegant boron-zinc exchange process startingfrom 1-oc tene developed by Knochel and co-workers, the 1,4-product 3.2i was26b

isol ated in 50% yield and with an e.e. of 72% (Eq. 3.4). In the determination of th eenantiomeric excess the same elution sequences was found for both enantiomers o f3.2h , 3.2i and 3.2a , indicating that again the S enantiomers were produced in excessemploying (+)-DAB as chiral ligand.

(3.4)

3.2i R = octyl, e.e. 72%(3.2h R = pentyl, e.e. 75%)

Ni(acac)2 (cat.)

Chalcone (3.1)CH3CN, - 30°C

(+)-DAB (cat.)Oct2Zn

1) HBEt22) Et2Zn( )

4*

OR

Ph Ph


15

The e.e.'s of 3.2h and 3.2i are slightly lower than in the case of diethylzinc, possiblydue to lower reactivity of the bulkier reagents. This can be explained by the tim edependenc y of the e.e. of the 1,4-product observed in the next Section. With th ela rger reagents the addition proceeds at a slower rate, resulting in longer reactio ntimes and lower e.e.'s ( vide infra). In Chapter 6 enantioselective conjugate additio nreact ions with functionalised dialkylzinc reagents, synthesised by utilising th ehydroboration-transmetalation protocol, will be described.26b

3.6 Mechanistic aspects

Time dependencyAlthough no quantitative measurements and detailed kinetic studies were performedand the mechanism of the enantioselective nickel catalysed conjugate addition i sstill unknown, two observations related to the mechanism will be discussed. Firstly,the dependence of the e.e. of 3.2a on reaction time and conversion of 3.1a wasstudied. At various intervals samples of the reaction mixture (Scheme 3.1, (-)-DAIBas chiral ligand) were taken and the conversion was determined by GC analysis .Ta ble 3.5 shows that after 90 min already 97% of 3.1a is converted. Th eenan tiomeric excess, determined by HPLC analysis, significantly decreased withtime (F igure 3.3). Whereas the product, isolated after 10 min, had an e.e. of 77 %29

the e.e. of 3.2a decreased to 68% after 90 min. Thus, at higher conversion of 3.1a ,product 3.2a with a lower e.e. is obtained.

0 20 40 60 80 10050

60

70

80

e.e.

of

3.2a

(%)

conversion (%)

Chapter 3

16

Table 3.5 Effect of reaction time on e.e. of product 3.2a.a

entry time (min) conv. (%) e.e. (%)b c

1 10 33 77 2 30 72 73 3 50 95 72 4 90 97 68 5 1260 > 99 65

a. Reaction conditions see Table 3.1. b. Conversion determined by GC analysis.c. See Table 3.1.

This remarkable time dependency might be explained by a mechanism proposed b yBolm. The organozinc reagent reduces nickel(II) to nickel(I) and nickel(0), o f4c

wh ich nickel(I) is most likely to be responsible for an efficient catalysis. Electro n30

transfer from nickel(I) to the substrate, followed by attack of the ketyl radical uponthe nickel(II) species thus generated forms a nickel(III) intermediate. Alky l31

transfer form zinc to the nickel center followed by reductive elimination gives th ezinc enolate of the 1,4-product and regenerates the catalytically active nickel(I )species (Scheme 3.6).

Ln*

Ph Ph

ONiEt ZnEt

Ph Ph

O

Ph Ph

O-.

Et2Zn

Ph Ph

O-Ni IIIL2*

Ph Ph

OZnEt

Et

nickel(I)

- nickel(I)Ph Ph

OEtH2O

Ni L2*II


17

Figure 3.3 Time dependency of the e.e. of product 3.2a.

The asymm etric induction is probably dictated in an enantioselective formation o fthe or ganonickel(III) intermediate followed by a stereoselective reductiv eel imination. After a certain time these enantioselective nickel(I) catalysts ar eprobably transformed into species which are still active, but which produce racemicmaterial. As a result, the overall e.e. of the product will decrease in time. It should beemphas ised that under the reaction conditions no racemisation of 3.2a occurs. In acontrol experiment the addition of another equivalent of diethylzinc and a portio nof chiral catalyst [Ni(acac) (7 mol%) and chiral ligand (16 mol%)] to the product in2

solution, did not affect the e.e. of 3.2a after 16 h at -30 EC.

Scheme 3.6 Proposed mechanism of the enantioselective nickel catalysedconjugate addition.

Asymmetric amplificationIn ord er to gain more insight in the nature of the catalytic active species, th erelationship between the e.e. of the ligand 3.5a and the e.e. of the product 3.2a wasdet ermined. Scalemic 3.5a with defined e.e. was employed in the nickel catalyse d32

co njugate addition of diethylzinc to chalcone ( 3.1a ). The results are summarised i nTable 3.6. As is clearly illustrated in Figure 3.4 a positive nonlinear relationship wasfound. The use of 3.5a with low e.e. resulted in the formation of 3.2a with an e.e .higher than expected on the basis of a linear relationship.

0 25 50 75 1000

20

40

60

80

e.e.

of

3.2a

(%)

e.e. of 3.5a (%)

7 mol% of Ni(acac) 2

1 mol% of Ni(acac) 2

Chapter 3

18

Table 3.6 Effect of e.e. of 3.5a on e.e. of 3.2a.a

e.e. of 3.2a (%) c

e.e. of 3.5a (%) 7 mol% of Ni(acac) 1 mol% of Ni(acac)b2 2

12.5 19 21 25 33 34 37.5 43 45 50 48 50 62.5 53 57 75 56 62 87.5 58 66 100 65 72

a. Reaction conditions see Table 3.1. b. Enantiomerically pure (-)-DAIB mixedwith racemic DAIB. c. See Table 3.1.

Figure 3.4 Effect of e.e. of 3.5a on e.e. of 3.2a.

Nonlinear relationships between the e.e.'s of chiral auxiliaries and e.e.'s of product swere described by Kagan and Agami and co-workers. Intensive investigations were33

car ried out in the enantioselective alkylation of aldehydes. These phenomen a1a, 34

have been interpreted in terms of differences in the chemical behaviour o fdiastereomeric mononuclear or dinuclear complexes. The asymmetri c35

(acac)NiO

N

(2R)-3.22

(2R, 2'R)-3.23

ON

NiO

N

(2S, 2'S)-3.23

N

ONi N

O

Ni(acac)O

N +

(2S)-3.22

(2S, 2'R)-3.23

NO

NiO

N


19

amplif ication in our system can be explained by the difference in chemica lproperties of diastereomeric complexes. However, it can not be excluded that largermultinuclear aggregates are involved in the catalytic process.It is reasonable to assume that two equivalents of scalemic DAIB replace bot hacetylacetonate anions from Ni(acac) forming diastereomeric mononuclear nicke l2

complexes of general structure NiL . Homochiral or heterochiral interaction between*2

the Ni( acac)L monomers (2 S)-3.22 and (2 R)-3.22 lead to the chiral dimeri c*

complexes, (2S, 2'S)-3.23 and (2 R, 2'R)-3.23 , or the meso complex (2 S, 2'R)-3.23(Sche me 3.7). It must be emphasised that only the cis complexes are shown ,however, trans coordination is also possible.

Scheme 3.7 Enantiomer recognition of (2S)-3.22 and (2R)-3.22 leading to possiblediastereomeric nickel complexes of scalemic DAIB (only cis-complexesare shown).

A positive nonlinear relationship has been explained by a greater stability of th emeso complex compared to the chiral diastereoisomers. Predominant reaction o f35

Chapter 3

20

di ethylzinc with the less stable optically active complex (2 S, 2'S)-3.23 (in case o fexcess of (2 S)-3.22 ) would lead to the formation of a homochiral active species. Theenantiomer o f the ligand, present in minor amount, is trapped in the more stabl emeso complex (2 S, 2'R)-3.23 , and becomes less available for catalytic conjugat eaddition.

3.7 Summary and concluding remarks

It has been demonstrated that the conjugate addition of diethylzinc to chalcone i seff ectively catalysed by chiral nickel complexes derived of tertiary amino alcohol swith enantioselectivities up to 85%. The highest e.e. values have been achieved with(- )-DAIB, (+)-DAB and the piperidyl substituted isoborneol ( 3.5g ). However, th econjugate addition of diethylzinc to cyclohexenone, catalysed by the chiral nicke lcomplex, proceeds without any enantioselectivity.Although the nature of the catalyst system is still unknown a mechanism has bee nproposed, mainly based on the remarkable time dependency of the e.e. of the 1,4 -product and the observed nonlinear effects. The intermediate which could accoun tfor the enantioselective transfer is still not clear, unfortunately.In order to develop a catalytic system capable of enantioselective conjugat eaddi tion to cyclic and acyclic substrates, tri- and tetradentate ligands, based on th esame $-amino alcohols, will be synthesised and examined in the nickel catalyse daddition.

3.8 Experimental section

GeneralAll reaction mixtures were stirred magnetically and performed without excluding airand moi sture unless stated otherwise. All conjugate addition reactions wit horganometallic reagents were performed under nitrogen. Tetrahydrofuran (THF) wasfreshl y distilled from Na/benzophenone and stored under nitrogen. Dichloro -methane (CH Cl ), diethyl ether, and hexane were distilled from P O and stored over2 2 2 5

4Å molecular sieves. Methanol and ethanol were distilled from Mg and stored ove r3Å molecular sieves. All nitriles, EtOAc and 2-propanol (p.a.) were purchased fro mJa nssen (now Acros) and were used without purification. H and C NMR spectr a1 13

were recorded on a Varian Gemini 200 (at 200 and 50.3 MHz respectively) or aVari an VXR-300 (300 and 75.5 MHz) spectrometer; solvent CDCl unless state d3


21

otherwis e. The chemical shifts are denoted in * units (ppm) relative to TMS ( * =0.00) for protons or CDCl (* = 76.91) for carbon atoms. Splitting patterns for H: s3

1

(singlet), bs (broad singlet), d (doublet), dd (double doublet), t (triplet), q (quartet) ,m (multiplet) and for C determined with the APT pulse sequence: q (quartet, CH ), t13

3

(triplet, CH ), d (doublet, CH), s (singlet, C). Optical rotations were measured on a2

Perkin -Elmer 241 MC (at RT). High Resolution Mass Spectra (HRMS) wer eobtained on a AEI MS-902 mass spectrometer by Mr. A. Kiewiet. GC analysis wa scarri ed out using a Hewlett-Packard 5890 II gas chromatograph (column: HP-1 )equipped with a Hewlett-Packard series II integrator. HPLC analysis was performedon a Waters 480 with a LC spectrophotometer or a Waters 600E system controlle rwith a Waters 991 photodiode array detector; Millenium 2010 as software .TM

Element al analyses were performed in the Microanalytical Department of thi slaboratory by Mr. H. Draaijer, Mr. J. Ebels and Mr. J. Hommes.

MaterialsTh e following compounds were commercially available and used withou tpurification: 1,3-diphenyl-2-propen-1-one (chalcone, 3.1a ; Janssen), substitute dchalcones 3.1b-f (Lancaster), nickel(II)acetylacetonate (anhydr.; Aldrich), NiBr 2

(anhydr.; Aldrich), diethylzinc (1 M in hexane; Aldrich, 15 wt % in hexane; Janssen,1.1 M in toluene; Aldrich), dimethylzinc (2 M in toluene; Aldrich), ( S)-(-)-1,1'-bi-2-na phthol (3.6 ; Syncom BV), (2 S,2'S)-2-(hydroxymethyl)-1-[(1-methylpyrrolidin -2yl) methyl]pyrrolidine (3.7 ; Merck), quinine ( 3.14 ; Janssen), 2.2'-bipyridin e(Merck), MeI (Merck) and 1, n-dibromoalkanes (Janssen).Cis-exo-3-aminoisoborneol (3.5b ), cis-exo-N-monomethyl-3-aminoisoborneo l10

((+)-MAIB) (3.5c ), (+)-diphenyl(1-methylpyrrolidin-2-yl)-methanol ((+)-DPMPM)22

(3 .8), ß-hydroxysulfonamides 3.9-3.11 mixed crown ethers 3.12 and 3.13 , cis-12 7 8

endo-3-aminoborneol (3.16b ) and trans-3-endo-aminoisoborneol ( 3.17b ) wer e19 10

prepared according to published procedures.

(-)-Cis-exo-N,N-dimethyl-3-aminoisoborneol (DAIB) (3.5a )Amino alcohol 3.5b (4.00 g, 23.6 mmol), MeI (7.25 g, 51.0 mmol), NaOH (3.0 g, 75mmol ) and 1 ml of H O were successively added to 30 ml of diethyl ether. Th e2

mixture was stirred for 16 h and an additional 2.00 g (14.0 mmol) of MeI was added.After another 24 h the mixture was poured into 50 ml of H O. The two layers wer e2

separated and the water phase was extracted with diethyl ether (2 x 30 ml). Th ecombined organic layers were washed with brine (50 ml) and dried (Na SO ).2 4

Filtrat ion and evaporation of the solvent gave 3.69 g of a crude oil. This oil wa spurified by bulb-to-bulb distillation (80-90 EC, 0.4 mmHg), column chromatography

N NPN ON

PNN

NOH

CDCl3, RT(3.5)

3.24

Chapter 3

22

(SiO , CH Cl /methanol (4:1)) and another bulb-to-bulb distillation (75 EC, 0.0 52 2 2

mmHg ) to provide pure 3.5a as a colourless oil (2.10 g, 10.6 mmol, 45%). [ "] -2 0D

14.5E (c 0.93, C H OH) [Lit. ["] -14.7 E (c 4.58, C H OH)] H NMR * 0.75 (s, 3H),2 5 D 2 518 2 8 1

0.84-0. 99 (m, 2H), 0.94 (s, 3H), 1.04 (s, 3H), 1.30-1.45 (m, 1H), 1.61-1.75 (m, 1H) ,1.93 (d, J = 4.7 Hz, 1H), 2.20 (d, J = 7.1 Hz, 1H), 2.26 (s, 2 x 3H), 3.40 (d, 7.1 Hz ,1H), 4.1-4.4 (bs, 1H). C NMR * 11.46 (q), 20.76 (q), 22.06 (q), 27.86 (t), 32.24 (t),13

46.30 (s) , 47.03 (d), 49.06 (s), 74.13 (d), 78.73 (d), N-CH was not observed in C313

NMR, due to unknown reasons. HRMS calcd for C H NO: 197.178, found 197.178.12 23

Racemic DAIB was obtained in a similar manner. The enantiomeric purity of 3.5awas determined by derivatisation with a chiral trivalent diazaphospholidine ( 3.24 )in CDCl (Eq. 3.5). With racemic DAIB two separated signals in the P NM R3

23 31

spectrum were observed, whereas in the case of (-)-DAIB one signal was absent (e.e.> 98%).

(+)-Cis-exo-N-ethyl-3-aminoisoborneol (3.5d )A mixture of 3.5b (1.69 g, 10 mmol), ethyliodide (3.95 g, 25 mmol) and K CO (3.462 3

g, 25 mmol) in 30 ml of ethanol was stirred and refluxed for 5 days. The mixture waspoured into 100 ml of H O and extracted with diethyl ether (3 x 30 ml). Th e2

combined organic layers were washed with brine (50 ml) and dried (Na SO ).2 4

Fil tration and evaporation of the solvent gave a crude oil. After bulb-to-bul bdistillation (100 EC, 0.1 mmHg), column chromatography (SiO , CH Cl /methano l2 2 2

(4:1)) an d extraction with 1 N NaOH, to remove silicium compounds, 3.5d wasisolated as an colourless oil (40%). [ "] + 6.4 E (c 0.5, C H OH) H NM R2 0 1

D 2 5

(CDCl /CD OD) * 0.73 (s, 3H), 0.86 (s, 3H), 0.90-1.00 (m, 2H), 0.97 (s, 3H), 1.15 (t, J3 3

= 7 .2 Hz, 3H), 1.33-1.47 (m, 1H), 1.58-1.73 (m, 1H), 1.80 (d, J = 4.6 Hz, 1H), 2.68 -2.92 (m, 3H), 3.53 (d, J = 7.4 Hz, 1H). C NMR (CDCl /CD OD) * 10.93 (q), 13.0 113

3 3

(q), 20.69 (q), 21.48 (q), 26.82 (t), 32.37 (t), 44.14 (t), 46.59 (s), 48.84 (s), 49.25 (d),


23

64.99 (d), 77.56 (d). HRMS calcd for C H NO: 197.178, found 197.178.12 23

General procedure for the N,N-dialkylation of primary amino alcohol 3.5bA mixture of 3.5b (1.69 g, 10 mmol), 1, n-dibromoalkane (11 mmol) and K CO (3.462 3

g, 25 mmol) in 30 ml of ethanol was stirred and refluxed for 16 h. The mixture wa spoured into 100 ml of H O and extracted with diethyl ether (3 x 30 ml). Th e2

combined organic layers were washed with brine (50 ml) and dried (Na SO ).2 4

Filtrat ion and evaporation of the solvent gave a crude oil. This oil was purified b ybul b-to-bulb distillation followed by column chromatography (SiO ,2CH Cl /methanol (4:1)). Yields 40-50% (colourless oils). The physical data of th e2 2

compounds are as follows:(-)-Cis-exo-3-(1-azetidinyl)isoborneol (3.5e )["] - 9.2E (c 1.0, CH Cl ) H NMR * 0.68 (s, 3H), 0.85 (s, 3H), 0.88-0.93 (m, 2H) ,2 0 1

D 2 2

0.99 (s, 3H), 1.30-1.45 (m, 1H), 1.53-1.72 (m, 1H), 1.64 (bs, 1H), 2.00-2.15 (quintet,J = 7.1 Hz, 2H), 2.55 (d, J = 7.1 Hz, 1H), 3.17-3.54 (m, 4H), 3.35 (d, J = 7.1 Hz, 1H).

C NMR * 11 .08 (q), 17.33 (t), 20.28 (q), 21.80 (q), 26.88 (t), 32.46 (t), 46.02 (s) ,13

46.68 (d), 4 9.13 (s), 55.63 (t), 74.89 (d), 78.67 (d). HRMS calcd for C H NO:13 23

209.178, found 209.178. Racemic 3.5e was obtained in a similar manner. (-)-Cis-exo-3-( 1-azetidinyl)isoborneol (3.5e ) was enantiomerically pure (e.e. > 98%), a sconfirmed by P NMR with a chiral derivatising agent (see also Eq. 3.5). 31 23

(+)-Cis-exo-3-(1-pyrrolidinyl)isoborneol (3.5f )["] + 14.8 E (c 0.4, CH Cl ) H NMR (CD OD) * 0.83 (s, 3H), 0.95 (s, 3H), 1.02-1.082 0 1

D 2 2 3

(m, 2H) , 1.12 (s, 3H), 1.45-1.55 (m, 1H), 1.67-1.95 (m and bs, 5H), 2.02 (d, J = 4. 7Hz, 1H), 2.66-3.07 (m, 5H), 3.54 (d, J = 7.3 Hz, 1H). C NMR (CD OD) * 11.40 (q),13

3

20.92 (q), 21.87 (q), 23.86 (t), 27.91 (t), 32.71 (t), 47.23 (s), 49.34 (d), 50.31 (s) ,56.12 (t), 74.84 (d), 78.87 (d). HRMS calcd for C H NO: 223.194, found 223.194.14 25

(+)-Cis-exo-3-(1-piperidinyl)isoborneol (3.5g )A mixtu re of 3.5b (1.69 g, 10 mmol), 1,5-dibromopentane (2.53 g, 11 mmol) an dK CO (3.46 g, 25 mmol) in 100 ml of acetonitrile was stirred and refluxed for 42 3

days. The mixture was poured into 100 ml of H O and extracted with diethyl ether (32

x 50 ml). The combined organic layers were washed with brine (50 ml) and drie d(Na SO ). Filtration and evaporation of the solvent gave crude 3.5g , which wa s2 4

purif ied by two successive column chromatography separations (SiO ,2CH Cl /methanol (10:1)). Yield 80% (colourless oil). ["] + 2.8 E (c 0.95, CH Cl ) H2 2 D 2 2

2 0 1

NMR * 0.76 (s, 3H), 0.95 (s, 3H), 0.89-0.98 (m, 1H), 1.00 (s, 3H), 1.25-1.55 (m an dbs, 9H), 1.80-2.05 (bs, 1H), 2.01 (d, J = 4.7 Hz, 1H), 2.05-2.30 (bs, 1H), 2.34 (d, J =7.0 Hz, 1H), 2.70-2.90 (bs, 1H), 2.90-3.10 (bs, 1H), 3.45 (d, J = 7.0 Hz, 1H). H NMR13

Chapter 3

24

* 11.43 (q), 20.85 (q), 22.06 (q), 24.34 (t), 26.05 (broad, t), 27.86 (t), 32.18 (t), 45.32(d), 46.52 (s), 48.90 (s), 50.63 (broad, t), 56.95 (broad, t), 72.86 (d), 78.48 (d) .HRMS calcd for C H NO: 237.209, found 237.209.15 27

(+)-Cis-endo-N,N-dimethyl-3-aminoborneol (DAB) (3.16a )(+)-DAB was obtained as a colourless oil. (Prepared from optically active activ e3.16b by the same procedure as described for (-)-DAIB.) [ "] + 53.2 E (c 0.9 ,2 0

D

CH Cl ) H NMR * 0.86 (s, 3H), 0.88 (s, 6H), 1.14-1.24 (m, 1H), 1.43-1.51 (m, 2H) ,2 21

1.66 (t, J = 3.8 Hz, 1H), 1.76-1.89 (m, 1H), 2.20 (s, 6H), 2.40 (dd, J = 8.5 Hz, J = 3.8Hz, 1H), 3.05-3.43 (bs, 1H), 3.63 (d, J = 8.5 Hz, 1H). C NMR * 14.41 (q), 18.85 (q),13

19. 01 (t), 20.03 (q), 26.40 (t), 45.13 (q), 46.89 (s), 48.46 (d), 50.27 (s), 66.07 (d) ,73.94 (d). HRMS calcd for C H NO: 197.178, found 197.178.12 23

(+)-Trans-N,N-dimethyl-3-aminoisoborneol (3.17a )Th is colourless compound was prepared from optically active active 3.17b by th esame pr ocedure as described for (-)-DAIB and solidified upon standing. [ "] +2 0

D

18.2E (c 1.0, CH Cl ) H NMR * 0.84 (s, 3H), 0.86 (s, 3H), 1.05 (s, 3H), 1.01-1.14 (m,2 21

1H), 1.45-1.65 (m, 3H), 1.77 (t, J = 4.1 Hz, 1H), 2.22 (s, 6H), 2.25-2.32 (m, 1H), 3.29(d, J = 3.0 H z, 1H). C NMR * 11.43 (q), 19.24 (t), 19.77 (q), 20.90 (q), 34.45 (t) ,13

44.58 (q), 47.03 (s), 48.21 (d), 49.66 (s), 76.90 (d), 84.85 (d). HRMS calcd fo rC H NO: 197.178, found 197.178.12 23

Conjugate addition of diethylzinc to chalcone (3.1a) using in situ prepared chiralnickel complexes; general procedureThis procedure is typical for all ligands. A solution of Ni(acac) (18 mg, 0.07 mmol)2

and 3.5a (32 mg, 0.16 mmol) in 2 ml of acetonitrile was stirred and refluxed for 1 hun der nitrogen. The solution was cooled to room temperature and 1,3-diphenyl-2 -propen-1-one (3.1a , 208 mg, 1.0 mmol) was added. The mixture was cooled to -35 ECand 1.5 ml of diethylzinc in hexane (1 M, 1.5 mmol) was added. The colour changedimmediate ly from bright green to dark brown red. Stirring was continued at -30 ECfor 16 h. The mixture was poured into 3 M aqueous HCl (15 ml) and extracted wit hCH Cl (3 x 20 ml). The combined organic layers were washed with brine (25 ml )2 2

and dried (MgSO ). Filtration and evaporation of the solvent gave crude 3.2a , which4

was purified by chromatography with a short column (SiO , CH Cl ) resulting in a2 2 2

colourless oil which solidified upon standing.Yields are shown in Tables 3.1-3.4. The e.e.'s were determined by HPLC analysis ;Da icel (Chiralcel OD), 0.25% iPrOH in hexane, flow rate 1.0 ml/min, UV detecto r(254 nm ); retention times ( S)-3.2a 16.3 min; ( R)-3.2a 19.0 min. Rac-3.2a gave tw o


25

base line sepa rated signals. H NMR * 0.85 (t, J = 7.3 Hz, 3H), 1.60-1.84 (m, 2H) ,1

3.24-3.33 (m, 3H), 7.18-7.33 (m, 5H), 7.41-7.60 (m, 3H), 7.92-7.96 (m, 2H). C NMR13

* 12.12 (q), 29.23 (t), 43.00 (d), 45.60 (t), 126.27 (d), 127.65 (d), 128.05 (d), 128.40(d), 128.52 (d), 132.90 (d), 137.25 (s), 144.67 (s), 199.19 (s). See also references 3and 4.

Conjugate addition of diethylzinc to the substituted chalcones 3.1b-e employingNi(acac) and (-)-DAIB2

The subs tituted chalcones 3.1b-e were alkylated as described in the precedin gproce dure. H NMR and C NMR data of the 1,4-products were in good agreemen t1 13

with the data found in the literature. E.e.'s are given in Table 3.4, entries 1-5.4c

Conjugate addition of diethylzinc to 3.1a using Ni(acac) , (-)-DAIB and additional2

achiral ligandAccording to the procedure described above the chiral catalyst was prepared in situin refluxing acetonitrile. The solution was cooled to room temperature and 3.1a (208mg, 1.0 mmol) and 2,2'-bipyridine (11 mg, 0.07 mmol) were added. Following th esame procedure as described above gave 3.2a (82%). E.e. values of 3.2a are given inTable 3.2, entries 2-5, and 7.

Time dependency of the conjugate additionThe general procedure as described above was applied and at various time intervalssamples of the solution (0.1 ml) were taken and quenched with 1 ml of 3 N aqueousHCl. Af ter extraction with 1 ml of CH Cl the organic layer was analysed by GC .2 2

Re tention times (oven temperature 225 EC, flow 101.8 ml/min He): 1,3 -diphenylpentan-1-one (3.2a ), 7.35 min; 1,3-diphenylpropenone ( 3.1a ), 8.49 min .Th e e.e. was determined by HPLC analysis. Reaction times, conversions and e.e .values are given in Table 3.5.

Conjugate addition; variation of solvent and temperatureAccordi ng to the general procedure, the chiral catalyst was prepared in situ fromNi(acac) and the chiral ligand in 2 ml of refluxing butyronitrile, propionitrile o r2

isob utyronitrile. At room temperature the substrate 3.1a and subsequently at -55 EC1.5 ml of diethylzinc (1 M in hexane) were added. After 16 h (temperature -50 EC)the conversion (> 90%) was determined by GC analysis (see previous procedure )and the pr oduct was isolated following the general work up procedure. The e.e. o f3.2a was d etermined by HPLC analysis. E.e. values of 3.2a are given in Table 3.4 ,entries 6-10.

Chapter 3

26

1,3-Diphenylbutan-1-one (3.2g )The gener al procedure was applied to create the chiral catalyst. The mixture wa scooled to -35 EC and 0.75 ml of dimethylzinc in toluene (2 M, 1.5 mmol) was added.After 16 h the usual workup procedure was followed. H NMR and C NMR data of1 13

3.2g were in good agreement with the data found in the literature. 4c

1,3-Diphenyloctan-1-one (3.2h )With (+)- DAB the general procedure was applied to create the chiral catalyst .Chalco ne (208 mg, 1.0 mmol) was added to the chiral catalyst solution at roo mte mperature, the mixture was cooled to -45 EC and dipentylzinc (2 mmol) wa s27

added. Stirring was continued at -25 EC for 16 h. The mixture was poured into 15 mlof 1 M aq ueous HCl and extracted with diethyl ether (3 x 20 ml). The combine dorgani c layers were washed with brine (25 ml) and dried (MgSO ). Filtration an d4

evaporati on of the solvent gave crude 3.2h , which was purified by colum nchrom atography (SiO , hexanes/EtOAc (10:1)) resulting in a colouless oil whic h2

solidifie d upon standing. Yield 70%. The e.e. was determined by HPLC analysis ;Da icel (Chiralcel OD), 0.25% iPrOH in hexane, flow rate 1.0 ml/min, UV detecto r(254 nm); retention times ( S)-3.2h 10.8 min; ( R)-3.2h 14.7 min. Rac-3.2h gave tw obase line s eparated signals. H NMR * 0.72-0.82 (m, 3H), 1.14-1.19 (m, 6H), 1.55 -1

1.63 (m, 2H), 3.15-3.28 (m, 3H), 6.98-7.42 (m, 8H), 7.80-7.85 (m, 2H). C NMR (no13

ass ignment) * 14.07, 22.53, 27.19, 31.80, 36.32, 41.32, 46.03, 126.25, 127.59 ,128 .08, 128.43, 128.54, 132.91, 137.30, 145.04, 199.24. HRMS calcd for C H O:20 24

280.183, found 280.183.

1,3-Diphenyl-1-undecanone (3.2i )Wit h (+)-DAB the same procedure as described above was applied to create th ech iral catalyst. The solution was cooled to room temperature and 3.1a (416 mg, 2. 0mmol) was added. The mixture was cooled to -20 EC and dioctylzinc in toluen e26b

(1.5 M, 3 m mol) was added. Stirring was continued at -15 EC for 13 h. Th econvers ion (~75%) was determined by GC analysis (see procedure above) an dstir ring was continued at room temperature for 1 day. The mixture was poured int o15 ml of 1 M aqueous HCl and extracted with diethyl ether (3 x 20 ml). Th eco mbined organic layers were washed with brine (25 ml) and dried (MgSO ).4

Filtration and evaporation of the solvent gave crude 3.2i , which was purified b ycolumn chromatography (SiO , hexanes/diethylether (30:1)) resulting in a colouless2

oil which solidified upon standing. Yield 50%. The e.e. was determined by HPL Canalys is; Daicel (Chiralcel OD), 1.0% iPrOH in hexane, flow rate 1.0 ml/min, U Vde tector (235.5 nm); retention times ( S)-3.2i 6.97 min; ( R)-3.2i 8.02 min. Rac-3.2i


27

1. See Section 1.2 for a short review of successful catalytic enantioselective carbon-carbon bon dform ing reactions. For reviews on catalytic asymmetric diethylzinc addition to aldehydes, see: a )Noyori, R.; Kitamura, M. Angew. Chem. Int. Ed. Engl. 1991 , 30, 49. b) Soai, K.; Niwa, S. Chem. Rev.1992 , 92, 833.

2. a) Greene, A.E.; Lansard, J.P.; Luche, J.L.; Petrier, C. J. Org. Chem. 1984 , 49, 931. b) Petrier, C.; deSouza Barbosa, J.C.; Dupuy, C.; Luche, J.-L. J. Org. Chem. 1985 , 50, 5761.

3. a) Soai, K.; Hayasaka, T.; Ugajin, S.; Yokoyama, S. Chem. Lett. 1988 , 1571. b) Soai, K.; Yokoyama,S.; Hayasaka, T.; Ebihara. K. J. Org. Chem. 1988 , 53, 4148. c) Soai, K.; Hayasaka, T.; Ugajin, S. J.Chem. Soc., Chem. Commun. 1989 , 516. d) Soai, K.; Okudo, M.; Okamoto, M. Tetrahedron Lett.1991 , 32, 95 (without nickel or any other transition metal).

4. a) Bolm, C.; Ewald, M. Tetrahedron Lett. 1990 , 31, 5011. b) Bolm, C. Tetrahedron: Asymmetry1991 , 2, 701. c) Bolm, C.; Ewald, M. Felder, M. Chem. Ber. 1992 , 125, 1205.

gave two base line separated signals. H NMR * 0.95 (t, J = 6.78 Hz, 3H), 1.21-1.371

(m, 12H), 1.71-1.82 (m, 2H), 3.31-3.42 (m, 3H), 7.24-7.61 (m, 8H), 7.97-8.00 (m ,2H). C NMR * 13.99 (q), 22.53 (t), 27.37 (t), 29.13 (t), 29.31 (t) 29.46 (t) 31.71 (t),13

36.22 (t), 41.19 (d), 45.87 (t), 126.08 (d), 127.44 (d), 127.91 (d), 128.26 (d), 128.36(d), 132.73 (d), 137.15 (s), 144.88 (s), 199.04 (s). HRMS calcd for C H O: 322.230,23 30

found 322.230.

Asymmetric amplificationThe ena ntiomeric excess of (-)-DAIB was adjusted by mixing appropriate amount sof enantiomeric pure (-)-DAIB and rac-DAIB to give 32 mg (0.16 mmol) of scalemicligan d, which was added to a solution of Ni(acac) (0.07 or 0.01 mmol) i n2

acetonitrile. The general procedure was followed and after 16 h of reaction time theconversion (> 95%) was determined by GC analysis (see previous procedure) an dthe p roduct was isolated following the general work up procedure. The e.e. of 3.2awas determi ned by HPLC analysis. E.e. values of 3.5c and 3.2a are given in Tabl e3.6.

Acknowledgements

Drs A. van Oeveren, J. Knol, and H.T. Stock are acknowledged for the synthesis o fcompound s 3.8 , 3.9-3.11 , and 3.12-3.13 , respectively. Prof. P. Knochel, Universit yof Marburg, is acknowledged, for the opportunity given to perform experimenta lwork in his laboratories. Mr. M. Suijkerbuijk is thanked for assistance with the manye.e. determinations of 3.2 .

3. 9 References and notes

Chapter 3

28


6. De Vries, A.H.M.; Jansen, J.F.G.A.; Feringa, B. L. Tetrahedron 1994 , 50, 4479.

7. Knol, J. Chiral Lewis Acid Catalyzed Diels-Alder Reactions, Ph.D. Thesis, University of Groningen,1996 .

8. Stock, H .T. Chiral Thiocrown Ethers, Synthesis and Application in Asymmetric Catalysis, Ph.D.Thesis, University of Groningen, 1994 .

9. ( - ) -DAIB was synthesised according to a published procedure and had the same optical purit y10

( ["] -9.77E (c 0.52, C H OH)) as reported by Noyori and co-workers, where (-)-DAIB served a s2 0D 2 5

an excellent ligand in the enantioselective addition of dialkylzinc reagents to aldehydes. 11

10. Chittenden, R.A.; Cooper, G.H. J. Chem. Soc. C 1970 , 49.

11. Kitamura, M.; Suga, S.; Kawai, K.; Noyori, R. J. Am. Chem. Soc. 1986 , 108, 6071.

12. (+)-DPMPM was synthesised by A. van Oeveren according to a literature procedure: Soai, K. ;Ookawa, A.; Kaba, T.; Ogawa, K. J. Am. Chem. Soc. 1987 , 109, 7111.

13. Sacconi, L.; Mani, F.; Bencini, A. In Comprehensive Coordination Chemistry; Wilkinson, G. ;Gillard, R.D.; McLeverty, J.A., Eds.; Pergamon: Oxford, 1987 , Vol. 5, Chapter 50.

14. (+)-Camphor: world production 1,000 tons per annum, approximate price 10 $/kg. Sheldon, R. A .Chirotechnology, Marcel Dekker, Inc., New York, 1993 , p. 144. See also Merck Index, Budavari, S.ed., 11 ed. Merck 1990 , No. 1738, p. 261.th

15. Oppolzer, W. Tetrahedron 1987 , 43, 1969.

16. a) Foster, M.O.; Rao, K.A.N. J. Chem. Soc. 1926 , 2670. b) Liu, J.-H. Angew. Makromol. Chem. 1987 ,155, 83.

17. a) Beckett, A.H.; Lan, N.T.; McDonough, G.R. Tetrahedron 1969 , 25, 5693. b) Marsman, B.G. TheCatalytic Asymmetric Synthesis of Optically Active Epoxy Ketones, Ph.D. Thesis, University o fGroningen, 1981 .

18. Noyor i, R.; Suga, S.; Kawai, K.; Okada, S.; Kitamura, M.; Oguni, N.; Hayashi, M.; Kaneko, T. ;Matsuda, Y. J. Organomet. Chem. 1990 , 382, 19.

19. a) Duden, P.; Pritzkow, W. Ber. 1899 , 32, 1538. b) Van Tamelen, E.E.; Judd, C.I. J. Am. Chem. Soc.1958 , 80, 6305. c) Beckett, A.H.; Lan, N.T.; McDonough, G.R. Tetrahedron 1969 , 25, 5689.

20. Pauling, H. Helv. Chim. Acta 1975 , 58, 1781.

21. Tanaka, K.; Ushio, H.; Kawabata, Y.; Suzuki, H. J. Chem. Soc. Perkin Trans. 1 1991 , 1445. See alsoM'Boungou-M'Passi, A.; Hénin, F.; Muzart, J.; Pète, J.P. Bull. Soc. Chim. Fr. 1993 , 130, 214.

22. Daniel, A.; Pavia, A.A. Bull. Soc. Chim. 1971 , 1060.

23. a) Hulst, R. New Methods for the Enantiomeric Excess Determination Using NMR, Ph.D. Thesis ,University of Groningen, 1994 . b) Hulst, R.; de Vries, N.K.; Feringa, B.L. Tetrahedron: Asymmetry1994 , 5, 699.

24. For a discussion on the dramatic effect of adding alike ligands in asymmetric catalysis, see: Faller,J.W.; Parr, J. J. Am. Chem. Soc. 1993 , 115, 804.

25. a) Corma, A.; Iglesias, M.; Martín, V.; Rubio, J.; Sánchez, F. Tetrahedron: Asymmetry 1992 , 3, 845b) Corma, A.; Iglesias, M.; del Pino, C.; Sánchez, F. J. Organomet. Chem. 1992 , 431, 233.


29

26. a) Knochel, P.; Singer, R.D. Chem. Rev. 1993 , 93, 2117. b) Langer, F.; Devasagayaraj, A.; Chavant,P.-Y.; Knochel, P. Synlett, 1994 , 410.

27. a) Seebach, D.; Behrendt, L.; Felix, D. Angew. Chem. Int. Ed. 1991 , 30, 1008. b) Weber, B.; Seebach,D. Angew. Chem. Int. Ed. 1992 , 31, 84. c) Von dem Bussche-Hünnefeld, J. L.; Seebach, D .Tetrahedron 1992 , 48, 5719.

28. For a disc ussion about ligand-accelerated asymmetric catalysis, see: Berrisford, D.J.; Bolm, C. ;Sharpless, K.B. Angew. Chem., Int. Ed. Engl. 1995 , 34, 1059.

29. Although we cannot exclude a minor effect of sampling on the enantioselectivity it should be notedthat less than 5% of the solution is removed and the concentration of reactants is not affected.

30. a) For a related mechanism proposed for nickel catalysed conjugate addition of organozirconiu mreagents to enones: Dayrit, F.M.; Gladkowski, D.E.; Schwartz, J. J. Am. Chem. Soc. 1980 , 102, 3976.b) The ni ckel(I)-nickel(III)-nickel(I) cycle is also proposed in the nickel-mediated aryl couplin greaction: Tsou, T.T.; Kochi, J.K. J. Am. Chem. Soc. 1979 , 101, 7547.

31. A similar electron transfer process has been proposed by House for the reaction of organocupratesto enones: House, H.O. Acc. Chem. Res. 1976 , 9, 59.

32. The term scalemic refers to unequal mixtures of enantiomers. Heathcock, C.H.; Finkelstein, B.L. ;Jarvi, E.T.; Radel, P.A.; Hadley, C.R. J. Org. Chem. 1988 , 53, 1922.

33. Puchot, C.; Samuel, O.; Dunach, E.; Zhao, S.; Agami, C.; Kagan, H.B. J. Am. Chem. Soc. 1986 , 108,2353.

34. Nonlinear effects: in ene reaction, Terada, M.; Mikami, T.; Nakai, T. J. Chem. Soc., Chem. Commun.1990 , 1623; in trimethylsilylcyanations, Hayashi, M.; Matsuda, T.; Oguni, N. J. Chem. Soc., Chem.Commun. 1990 , 1364; in stoichiometric conjugate addition of organocuprates to enones, Rossiter ,B.R.; Eguchi , M.; Miao, G.; Swingle, N.M.; Hernández, Vickers, D.; Fluckiger, E.; Patterson, G. ;Reddy, K.V. Tetrahedron 1993 , 49, 965.

35. See ref. 1a and also Wynberg, H.; Feringa, B.L. Tetrahedron, 1976 , 32, 2831.

1

Chapter 4

(+)-Camphor-Derived Tri- and Tetradentate Amino Alcohols; Synthesis and Application as Ligands in the

Nickel Catalysed Enantioselective Conjugate Addition of Diethylzinc

4.1 Introduction

In the p revious Chapter we have seen that the conjugate addition of diethylzinc t ochalcone is effectiv ely catalysed by chiral nickel complexes derived from (-)-DAIB and(+)- DAB. Although the nature of the catalytically active complex is unknown, amec hanism has been proposed which could account for the enantioselective alky ltransf er. Probably a nickel complex with two coordinated amino alcohol ligands i sinvolved in the rate determining step of the catalytic process. However, it should beemphasised th at participation of multinuclear aggregates, consisting of nickel and zinccenters and sever al chiral ligands, can not be excluded. Furthermore, the nickel / aminoalcohol catalysed conjugate addition of diethylzinc was only enantioselective fo racyclic enones i.e. chalcones. Alkyl transfer from diethylzinc to cyclohexenon eproc eeded smoothly, presumably via a kinetically favoured complex or aggregat eincapable of inducing enantioselectivity. This Chapter describes attempts to isolate a mononuclear nickel complex derived ofNi(acac) and two equivalents of (-)-DAIB (or (+)-DAB). Furthermore, to develop a2

catalyst capable of enantioselective conjugate addition of dialkylzinc reagents to cyclicand acyclic enones, several novel tri- and tetradentate tertiary amino alcohols, al lderived of (+)-camphor, were synthesised. Probably the rigidity of the in situ preparedcataly st will be enhanced with these multidentate ligands, minimalising the number ofpos sible aggregates and as a result inducing enantioselectivity for both types o fsubstrates.

Crystallisation experimentsSynthesis, isolation, and characterisation of cis-complex 4.2 (Eq. 4.1) or the transcoordinated comp lex by a X-ray structure determination should give information aboutthe comple x, generated in situ in the catalytic process. Therefore, a solution o fNi(aca c) and 2 equivalents of (-)-DAIB ( 4.1) in acetonitrile was heated at reflux for 22

h and cooled slowly in an attempt to obtain crystals suitable for X-ray analysis. N ocrystals wer e formed after 1 week at -20 EC and slow evaporation of the solvent did notfurni sh suitable material as well. Unfortunately, in toluene or with NiBr instead o f2

Ni(ac ac) , only green powders were collected. H NMR measurements on thes e21

4.1

NOH

2

4.2

NO Ni N

ONi(acac)2 (4.1)

Chapter 4

2

materials in CD CN revealed no changes in chemical shifts compared to the free ligand.3

Also the UV spectrum of the presumed Ni(acac) -ligand complex gave the sam e2

absorption max ima as observed for the ligand spectrum, indicating no ligand exchangeunder the se conditions. Probably deprotonation of the alcohol group of 4.1 is requiredto achieve compl exation of 4.1 to nickel. However, the use of Et N, NH , NaOH, or NaH3 3

did not furnish material suitable for crystal structure determination or Ni(acac) was2

isolated.

When the nickel catalysed conjugate addition reactions, described in Chapter 3, wereperformed, a slight excess of diethylzinc was used. Probably the chiral amino alcoholis deprotonated by diethylzinc forming a LZn-alkyl complex whereupon complexation1

to the nickel center and successive 1,4-addition can occur. Therefore, a clear solutionof (-)-D AIB (or stereoisomer (+)-DAB) in acetonitrile was treated with 1 equivalent ofdiethylzinc (1.1M in toluene) resulting in a turbid white mixture, which was cooled1

with ice and quenched with 0.5 equivalent of Ni(acac) . Gas evolved and the mixture2

immediately tur ned red brown, indicating the formation of a nickel complex. The clear2

solution was allowed to stand for 3 days at room temperature, resulting in an orangepr ecipitate. The H NMR spectrum of the precipitate showed broad signals .1

Unfortunately, recrystallisation of the precipitate from toluene / acetonitrile did no tgive crystals suitable for X-ray analysis. Possibly other geometries of the prepare dnicke l-ligand complex ( trans coordination of two amino acohol ligands is als opossible) and larger aggregates are present as well.

2

4.1a R = H4.1b R = Me

4.3

(4.2)NOH

RH N

OH HON

alkyl bridge

BrCH2 CH2Br

K2CO3, ethanolreflux

HON

OHN

4.1b 4.3a

(4.3)2N

OHH

(+)-Camphor Derived Tri- and Tetradentate Amino Alcohols; Synthesis and Application

3

4.2 Synthesis of (+)-camphor-derived tri- and tetradentate amino alcohols

Tetradentate amino alcoholsIn order to minimalise the possible geometries of the in situ prepared nickel-ligan dcomplexes, we were interested in the development of alkyl bridged aminoisoborneolsfurni shing tetradentate ligands of type 4.3 (Eq. 4.2). Several attempts to synthesis ealkyl bridged aminoisoborneols by straightforward coupling of N-methyl-3 -aminoisoborneol ((+)-MAIB) ( 4.1b ) with 1,3-dibromopropane or 1,4-dibromobutanefailed, probably due to steric hindrance. Corresponding tertiary amino alcohols wereonly synthesised by using the primary amino alcohol and dibromoalkanes using thefav oured intramolecular second substitition (see also Section 3.3). Therefore, othe r

routes had to be developed. Ally lic and benzylic substrates undergo nucleophilic substitution especially easily ,owing to reso nance stabilisation of the transition state. Reaction of (+)-MAIB with the3

do uble allylic substrate ( E)-1,4-dibromo-2-butene under the conditions given i neq uation 4.3, furnished the coupled ligand 4.3a in a moderate isolated yield (42%) .Unfortunately, reduction of the double bond in compound 4.3a with hydrogen in thepresence of P d/C, resulted in a mixture of compounds. We were not able to isolated thedesired compound from this mixture.

Reaction of (+) -MAIB with the double benzylic substrate ","'-dibromo- m-xylene under

4.4

H

O O

H

4 NaBH3CN4 NaOAcethanol / H 2O

N

OH(4.4)

4.1a

2NH2

OH

Chapter 4

4

the same conditions as given in Eq. 4.3 also gave a coupled tetradentate product .Howev er, no pure material was obtained after attempts of purification (colum nchromatography and c rystallisation). Since these two bisallylic compounds did not givesat isfactory results and other bisallylic halides are not readily available, anothe rprocedure for achieving tetradentate ligands was desirable.Secondary and tertiar y amines can also be obtained via reductive amination of carbonylcompounds . At pH 6-7 the reduction of aldehydes and ketones with NaBH CN i s3

negligible whereas in this pH range the reduction of the iminium group ( i.e., >C=N R+2

or >C=N HR) proceeds smoothly resulting in the corresponding amine. Although the+ 4

presenc e of water hampers the initial iminium formation, a commercially availabl esolu tion of glutaric dialdehyde in water (25 wt%) was used in a double reductiv eamination with primary amine 4.1a (Eq. 4.4). Remarkably, only tertiary amine 4.4 was5

for med in a relatively good yield (50%). This indicates the large rate differenc ebetween the intramolecular iminium formation of a secondary amine and th eintermolecular iminium formation of a primary amine, in favour of the sterically morehindered intramolecular reaction.

Although this intramolecular reaction might be suppressed by using a large excess ofthe amine, it also suggested the possibility to use the sterically hindered secondary4

amine ( +)-MAIB ( 4.1b ). Reaction of glutaric dialdehyde with two equivalents of 4.1b ,however, f urnished the double oxazolidine 4.5 in a 60% isolated yield (Eq. 4.5). Againan intramolecular reaction, i.e. the nucleophilic addition of the alcohol moiety to theimi no intermediate, is much faster than a intermolecular reduction of the latte rcompound.

4.54.1b

ON

NO2

N

OH

H

H

O O

H

4 NaBH3CN4 NaOAcethanol / H2O

(4.5)


5

The acetal carbon-oxygen bond of an oxazolidine can be cleaved by reduction withLiAlH , leading to an N-substituted amino alcohol. However, treatment of compound4

6

4.5 with an excess of LiAlH in ether, dioxane, or THF resulted in unreacted starting4

material, a mixture of unidentified compounds, and products due to cleavage of th eacetal carbon-nitrogen bond, respectively.When (+)-MAIB was protected with trimethylsilyl chloride and coupled subsequentlywith glutaric dialdehyde under the conditions given in Eq. 4.5, again the doubl eoxazolidine 4.5 was detected in the crude product. From these results it was clear thatwe had to t urn to an alternative procedure to achieve tetradentate amino alcoho lligands. Properly speaking, the direct N-alkylation of secondary amino alcohols 4.1b or 4.8 isth e only versatile option and therefore this procedure was reinvestigated. Th enucleoph ilic substitution of the secondary amine (+)-MAIB ( 4.1b ) to alkyl halide sother than methyl iodide does not proceed in refluxing ethanol, ethyl acetate, or DMFunder basic conditions (see also Section 3.3). More promising results were achievedwhen stereoisom er N-methyl-3-aminoborneol ((+)-MAB, 4.8), synthesised and purifiedaccor ding to a literature procedure (Scheme 4.1), was used as nucleophile in th e7

coupling reaction w ith dibromoalkanes. Apparently, the endo stereoisomer is somewhatless sterica lly demanding and therefore better accessible for alkyl halides. In refluxingacetonitrile and in the presence of one equivalent K CO , tetradentate amino alcohols2 3

4.9a and 4 .9b were synthesised and isolated in remarkable high yields (ca. 80% ,Schem e 4.1). This substitution reaction was not successful with 1,4-dibromobutane asalkylating reagent. P robably the nucleophilic substitution in the former cases is assistedby neighboring groups. Under these optimised conditions other sterically hindere d3

(intramo lecular) N,N-alkylations were performed in good yields as well (see fo rexample Section 3.3).

4.7 4.84.6

c

ba

HONH

ONH

OHO

NH2

HON

HON( )n

4.9a n = 14.9b n = 2

Chapter 4

6

Scheme 4.1 Synthesis of tetradentate amino alcohols 4.9. (a) diethyl carbonate,KCO, reflux (50%) (b) LiAlH , THF, reflux (95%) (c) 1,2-dibromoethane2 3 4

or 1,3-dibromopropane, K CO , acetonitrile, reflux (80%).2 3

With tetradentate ligand 4.9b the corresponding crystallisation experiments a sdescribed above for bidentate ligand 4.1 were performed. So, when a mixture of 4.9bin aceton itrile / toluene was treated with two equivalents of diethylzinc (1.1M i ntoluene) a clear yel low solution was obtained and gas evolved (probably ethane). Next,1

one equivalent of Ni(acac) was added resulting in a clear green solution. The colour2

of the solution changed slowly (in 1 h) to orange. Upon standing at room temperatureno solid material was obtained; slow diffusion of diethyl ether into the mixture did notgive satisfactory results either. All further attempts to isolate crystals suitable for X-rayanalysis were not successful, so far.

Tridentate amino alcoholsFor comparison tw o novel tridentate amino alcohols, both derived of (+)-camphor weresynt hesised. Since Tanaka and co-workers achieved very high e.e.'s (> 90%) in th eenantioselective conjugate addition of methyllithium to a cyclic enone mediated (orcatalysed) by a copper complex of methylpyrrole substituted aminoborneol 4.10 (see7, 8

Section 2.4), we were interested in the behaviour of this secondary amine and th ecorresponding tertiary amine compound in the nickel catalysed conjugate addition ofdiethylzinc. Therefore, compound 4.10 was synthesised according to the literatur e

4.114.10

HON

N

a

HONH2

HONH

N

4.6

b

4.124.8

(4.6)HO

N

NHONH K2CO3

acetonitrilereflux

NCl .HCl


7

procedure and subsequently N-methylated with an excess of methyl iodide (Scheme7

4.2).

Scheme 4.2 Synthesis of amino alcohols 4.10 and 4.11. (a) i. 1-methylpyrrole-2-carboxaldehyde, Na SO , dichloromethane. ii. LiAlH , diethyl ether (45%)2 4 4

(b) excess MeI, NaOH, diethyl ether, water (83%).

Due to the aro maticity of pyrrole, the electron pair on the nitrogen of the pyrrole groupis hardly accessible for metal complexation. However, the synthesis of the bette rcoordinating pyridine substituted analogous amino alcohol 4.12 , by employin gpyridine-2-carboxaldehyde, failed (conditions as in Scheme 4.2); the pyridine groupwas N-methylated as well in the last step. Fortunately, a second route to the pyridinesubstituted amino alcohol using secondary amine 4.8 and 2-picolyl chlorid ehydrochloride furnished the desired tridentate ligand 4.12 in 75% yield (Eq. 4.6).

The enantiomeric purity of all novel tri- and tetradentate ligands was not determined,but on the basis of the reaction conditions used no epimerisation (or racemisation) isex pected. Furthermore, (-)-DAIB ( 4.1) and compound 3.5e , synthesised vi aanalogueous procedures, proved to be enantiomerically pure, as described in Section3.3.

Chapter 4

8

4.3 (+)-Camphor-derived tri- and tetradentate $$-amino alcohols as chiralligands in the nickel catalysed addition of diethylzinc to chalcone orcyclohexenone

Chalcone as substrateWith the chiral ligands described in the previous Section, we examined the effect ofadd itional coordinating groups on the enantioselectivity in the nickel catalyse daddition of diethylzinc to chalcone. The results are summarised in Table 4.1. Ligand4.3a , containing a bridge with a double bond, gave 1,4-product 4.14 with 49% e.e. ,which is lower than found with two equivalents of the corresponding bidentate ligand4.1 (65% e.e., see Section 3.4). Possibly the trans olefinic bond prevents coordinationof bo th amino alcohol moieties in ligand 4.3a to the nickel center, resulting in les senantios elective catalysis. In Chapter 3 it was already shown that at least tw oequivalents of bidentate amino alcohols are required for high enantioselectivities.

PhPh

O

4.13 4.14

CH3CN, -25°C

Ni(acac)2 (cat.)chiral ligand (cat.)

Ph Ph

O

*+ Et2Zn


9

Table 4.1 (+)-Camphor derived amino alcohol compounds as ligand in the nickelcatalysed enantioselective conjugate addition of diethylzinc to chalcone(4.13).a

entry chiral ligand (mol%) yield (%) e.e. (%) abs. conf.b c d

1 4.3a (8) 66 49 Re

2 4.5 (8) 75 2 Re

3 4.9a (8) 72 21 S 4 4.9b (8) 88 69 Se

5 4.10 (11) 75 57 S 6 4.11 (11) 91 80 S 7 4.11 (7) 78 65 S 8 4.11 (16) 83 83 S 9 4.12 (16) 67 - 0 10 4.16 (16) 74 - 0

a. Reactions at -25EC in 2 ml of acetonitrile using an in situ prepared catalyst from 7 mol% Ni(acac) and given2

amount of chiral ligand. 1.5 Equivalent of diethylzinc in toluene (1.1M) was used (unless stated otherwise).Reaction time 16 h. b. Isolated yield of the 1,4-product. c. Determined by HPLC analysis. (see Chapter 3) d.Comparison of retention times of 4.14 with known data gave the absolute configuration. e. Diethylzinc in9

hexane (0.9M) was used.

From the r esults found with the double oxazolidine 4.5 it is obvious that a free alcoholfunction must be present. Although diethylzinc is activated by the oxazolidine, n oenantiose lectivity was observed. With the alkyl bridged ligands 4.9a and 4.9balternating e.e. valu es for 4.14 were obtained, compared to the e.e. value found with twoequiv alents of the corresponding bidentate ligand 4.15 (82% e.e., Section 3.4) .Probably s omewhat different aggregates are formed in solution, especially with ligand4.9a , resulting in less enantioselective alkyl transfer. Compound 4.9 prohibits transcoordination of both amino alcohol moieties by the alkyl bridge, which may be anotherexplanation for the observed decrease in enantioselectivity.

4.164.15

HON

OH

N

Chapter 4

10

When the secondary amino alcohol 4.10 was employed as chiral ligand a remarkablehigh enantioselec tivity (57% e.e.) has been determined for 1,4-product 4.14 . In Chapter3 prim ary and secondary amino alcohols were also used as ligand, however, n oenantiose lectivity was observed. In spite of the proposed structure, responsible for thehighly enantioselective alkyl transfer found by Tanaka, where coordination of th e8

pyrrol e group is totally ignored, the role of the pyrrole group in this alkyl transfer cannot be excluded . However, it is possible that only steric effects of the pyrrole group areresponsible fo r the enantioselectivity found. Furthermore, it should be emphasised thatin these experiments the diethylzinc is added as solution in toluene instead of th eformerly used hexane solution. In comparison, when aminoborneol (+)-DAB wa semployed as chiral ligand, the change of co-solvent resulted in a minor enhancementof the e.e. found for 4.14 (82% vs. 87% e.e.).With t he tertiary amine ligand 4.11 , enantioselectivities for 4.14 were observed whichare com parable with those found for (+)-DAB ( 4.15 ) (entries 6-8). An interestin gfeature of 4.11 is that even with a ligand-to-nickel ratio of 1 a significant e.e. value of65% was found, indicating a role of the pyrrole moiety ( vide supra).Rather to our surprise the pyridine substituted ligand 4.12 furnished the 1,4-productwith no enantioselectivity at all (entry 9). Most probably, only the pyridine group isinvolved as a coordinating group in the in situ preparation of the catalyst, resulting ina complex or aggregate with the chiral backbone too far away from the active center.To compare this remarkable result, bidentate ligand 4.16 , synthesised by nucleophilicadditio n of monolithiated 2,6-lutidine to (+)-camphor, was tested as ligand in th e10

nickel cat alysed conjugate addition of diethylzinc to 4.13 . Again no enantioselectivitywas found wi th this pyridine substituted ligand (entry 10). Although ligand 4.16 is a (-pyridine alcohol instead of a $-amino alcohol as all other ligands, the pyridine entityseems to be detrimental to asymmetric induction in this conjugate addition reaction.All novel tri- and tetradentat eamino alcohols furnished the sameen antiomer of 4.14 in excess a scom pared to the results found fo rthe analogous bidentate ligand s(see Chapter 3), indicating that thedirecti on of asymmetric induction isnot influenced by introducin gadditional coordinating and / or sterically demanding entities.

Ni(acac)2 (cat.)L* (cat.)

acetonitrile-25°C

Et2Zn+

4.17

OO

4.18

*


11

Cyclohexenone as substrateAs shown above the tri- and tetradentate ligands were able to induce asymmetric ethyltransfer to ch alcone. In order to investigate whether this enantioselective alkyl transfercan also be achieved with cyclic substrates, compounds 4.9-4.12 and 4.16 wer eexamined as chiral ligands in the conjugate addition of diethylzinc to cyclohexenone(4.17 ). The results are summarised in Table 4.2.

Table 4.2 Amino alcohols 4.9-4.12 and 4.16 as chiral ligand in the Ni(acac)2

catalysed 1,4-addition of diethylzinc to cyclohexenone.a

entry chiral ligand (mol%) e.e. of 4.18 (%) abs. conf.b c

1 4.9a (8) - 0 - 2 4.9b (8) - 0 - 3 4.10 (16) 7 S 4 4.11 (16) 12 S 5 4.12 (16) - 0 - 6 4.16 (16) 20 R

a. Reactions at -25EC in 2 ml of acetonitrile using an in situ prepared catalyst from 7 mol% Ni(acac) and given2

amount o f chiral ligand. 1.5 Equivalent of diethylzinc in toluene (1.1M) was used. Reaction time 16 h .Conversion to the 1,4-product > 90% as determined by GC analysis. Isolated yields of 3-ethylcyclohexanone( 4.18 ) > 70% . b. Enantiomeric excess of 4.18 was determined by derivatisation with optically pure 1,2 -diphenylethylene diamine. c. Comparison of the optical rotation of 4.18 with known data gave the absolute11

configuration (see Chapter 5).

In all cases the 1,4-product 4.18 was isolated in good yields (> 70%). However, withtetrade ntate ligand 4.9 no enantioselectivity was observed, and tridentate ligands 4.10and 4.11 were not able to induce selectivity exceeding 12% e.e. (for ligand 4.11 ). Thepyridi ne substituted tridentate ligand 4.12 gave no enantioselectivity, wherea sbiden tate ligand 4.16 furnished 1,4-product 4.18 with an e.e. of 20%. Remarkably, the

Chapter 4

12

ligands 4.10 and 4.11 furnished the S enantiomer of 4.18 in slight excess. This is theenantiomer found a s product in the reaction with chalcone (s- cis enone), indicating thatwith cyc lohexenone (s- trans enone) another aggregate, present in solution, i sresponsible for the enantioselective alkyl transfer.


In this Chapter the synthesis of several novel tri- and tetradentate amino alcohols, allderived from (+)-camphor, have been described. The initial attempt to synthesise N-al kylated amino alcohols by reductive amination was hampered by intramolecula rreaction w ith the alcohol moiety. Although the N,N-dialkylations were hindered by thesterically demanding backbone, successful procedures for the synthesis of tertiar yamines were developed by using reactive alkyl halides in refluxing acetonitrile .Unfortu nately, crystallisation experiments of the complexes of these multidentat eligands and the corresponding bidentate ligands with nickel were unsuccessful so far.Catalytic e nantioselective conjugate additions of diethylzinc, employing the tetra- andtridentate amino alcohols were successful with chalcone as substrate. About the sameenantioselectivities were achieved compared to the corresponding bidentate ligands.In contrast to the results described in Chapter 3, the secondary amino alcohol 4.10 wassuccessful as a chiral ligand in the enantioselective conjugate addition of diethylzincto chalco ne. Apparently, sterically demanding substituents on the amine entity ar ecrucial for enantio selective catalytic conjugate addition reactions and the most excitingligand stil l has to be prepared. Remarkably, the tridentate ligand with a pyridin esubst ituent furnished the 1,4-product without any enantioselectivity, indicating th every specific ligand requirements.Unfortunately, the i nitial goal to develop a catalytic system, capable of enantioselectiveconjugate addition of diethylzinc to both cyclic and acyclic substrates, failed. Wit hcyclohexenone as substrate good yields of 1,4-product were isolated, however, e.e.'swere not exceeding 20%. Therefore other catalytic systems, derived from other metalsalts and chiral ligands, seem to have more potency.


For general remarks, see Section 3.8.

MaterialsTh e following compounds were commercially available and used without furthe r


13

purification: (E)-1,4-dibromo-2-butene (Aldrich), ","'-dibromo- m-xylene (Aldrich),glutari c dialdehyde (25 wt% in water, Aldrich), NaBH CN (Aldrich), 1-methylpyrrole-3

2-carboxaldehyde (Aldrich), 2-pyridinecarboxaldehyde (Aldrich), 2-picolyl chloridehydrochloride (Aldrich). For the synthesis of compounds 4.1 , 4.1a , 4.1b , 4.6 and 4.15see Chapter 3. (+)-Cis-endo-N-[(1-methylpyrrol-2-yl)methyl]-3-aminoborneol ( 4.10 )was synthesised according to a literature procedure. 7

For all other materials, see Section 3.8.

Attempted synthesis of N,N'-Bis[3-cis-exo-isoborneol]-N,N'-dimethyl-1,m-alkyldiamine (general procedure for 4.3)A mixture of cis-exo-N-monomethyl-3-aminoisoborneol ((+)-MAIB, 4.1b ) (0.367 g, 2.0mmol), 1,m-dibromoalkane (1.0 mmol) and K CO (0.276 g, 2.0 mmol) in 50 ml o f2 3

ethanol (ethyl acetate or DMF) was stirred and refluxed for 1 week. The mixture waspoured into 2 5 ml of H O and extracted with CH Cl (3 x 30 ml). The combined organic2 2 2

layers were washed with brine (50 ml) and dried (Na SO ). Filtration and evaporation2 4

of the s olvent gave a crude oil that showed several spots on TLC (SiO ,2CH Cl /methanol (4:1)). We were not able to isolate the desired product from thi s2 2

mixture.

(-)-(E)-N,N'-Bis[3-cis-exo-isoborneol]-N,N'-dimethyl-1,4-diamino-2-butene (4.3a )A mixture of cis-exo-N-monomethyl-3-aminoisoborneol ((+)-MAIB) ( 4.1b ) (0.446 g,2.43 mmol), ( E)-1,4-dibromo-2-butene (0.260 g, 1.22 mmol) and K CO (0.442 g, 3.22 3

mmol) in 25 ml of ethanol was stirred and refluxed for 3 days. The mixture was pouredinto 25 ml of H O and extracted with CH Cl (3 x 50 ml). The combined organic layers2 2 2

were washed with brine (25 ml) and dried (Na SO ). Filtration and evaporation of the2 4

solvent gave a crude orange solid, which was purified by column chromatograph y(SiO , CH Cl /methanol (10:1)). Yield 42%. [ "] -11.7 E (c 1.0, CH Cl ). H NMR * 0.772 2 2 D 2 2

2 0 1

(s, 6H), 0.95 (s, 6H), 0.99-1.04 (m, 2H), 1.05 (s, 6H), 1.35-1.50 (m, 2H), 1.63-1.79 (m,2H), 2.00 (d, J = 4.6 Hz, 2H), 2.27 (s, 6H), 2.44 (d, J = 7.0 Hz, 2H), 2.8-3.2 (broa dsignal, 2 H), 3.15 (bs, 1H), 3.21 (bs, 1H), 3.45 (d, J = 7.0 Hz, 2H), 3.7-4.0 (broad signal,2H), 5.62-5.68 (m, 2H). C NMR * 11.43 (q), 20.98 (q), 22.02 (q), 27.87 (t), 32.15 (t),13

46.49 (s), 46. 72 (d) 49.13 (s), 72.50 (d), 78.69 (d), 130.05 (d). N-alkyl signals were notobserved in C NMR due to unknown reasons. HRMS calcd for C H N O : 418.356,13

26 46 2 2

found 418.356.

(+)-Cis-exo-3-(1-piperidinyl)isoborneol (4.4)A mixture of cis-exo-3-aminoisoborneol ( 4.1a ) (4.95 g, 29.2 mmol), glutaric dialdehyde(25 wt% in H O) (5.9 g, 14.6 mmol), NaOAc (4.8 g, 58.5 mmol) and NaBH CN (3.7 g,2 3

Chapter 4

14

58.5 m mol) in 100 ml of ethanol was stirred at ambient temperature for 9 days. Excessof aqueous HCl (3M) was added carefully to destroy excess NaBH CN. The aqueous3

layer was adjusted to pH 10 with saturated aqueous Na CO and extracted with CH Cl2 3 2 2

(2 x 125 ml) and ethyl acetate (1 x 100 ml). The combined organic layers were washedwith br ine (100 ml) and dried (Na SO ). Filtration and evaporation of the solvent gave2 4

a cr ude orange oil, which was purified by column chromatography (SiO ,2CH Cl /methanol (20:1)). Yield 1.73 g (50% compared to glutaric dialdehyde). Al l2 2

spectroscopic data were in good agreement with the given structure of 4.4 and similaras those described in Chapter 3.

(+)-1,3-Bis[5-methyl-1,2-cis-exo-bornane-3,5-oxazolidine-4-yl]propane (4.5)A mixture of (+)-MAIB) ( 4.1b ) (0.642 g, 3.5 mmol), glutaric dialdehyde (25 wt% inH O) (0 .726 g, 1.8 mmol), NaOAc (0.590 g, 7.2 mmol) and NaBH CN (0.452 g, 7. 22 3

mmo l) in 25 ml of ethanol was stirred at ambient temperature for 2 days. Excess o faqueous HCl (3 M) was added carefully to destroy excess NaBH CN. The aqueous layer3

was adjusted to pH 10 with saturated aqueous Na CO and extracted with CH Cl (3 x2 3 2 2

50 ml) and ethyl acetate (1 x 50 ml). The combined organic layers were washed withbrine (100 ml) and dried (Na SO ). Filtration and evaporation of the solvent gave a 2:12 4

mixture of product and (+)-MAIB (according to H NMR). Column chromatography1

(SiO , CH Cl /methanol (20:1)) afforded compound 4.5 as a white solid in 60% yield.2 2 2

["] +30.7 E (c 1.0, CH Cl ). H NMR * 0.79 (s, 6H), 0.81-0.88 (m, 2H), 0.96 (s, 6H),2 0 1D 2 2

1.18 (s, 6H), 1.35-1.74 (m, 14H), 2.26 (s, 6H), 2.35 (d, J = 7.7 Hz, 2H), 3.54-3.62 (m,2H), 3.68 (d, J = 7.7 Hz, 2H). C NMR * 10.95 (q), 20.26 (q), 22.09 (t), 22.79 (q), 25.3613

(t), 32.40 (t), 32.78 (t), 38.61 (q), 46.49 (s), 46.93 (d) 47.35 (s), 74.83 (d), 87.78 (d),97.48 (d). HRMS calcd for C H N O : 430.356, found 430.356.27 46 2 2

(+)-Cis-endo-N-monomethyl-3-aminoborneol ((+)-MAB) (4.8)This compound was synthesised according to a literature procedure (see also text) .7

Howeve r, in our hands the synthesis (under inert atmosphere) and purification o fcompound 4.7 was n ot that successful, yield ca. 50% (literature 85%). Reduction of 4.77

with LiAlH in THF gave 4.8 in 96% yield. H NMR and C NMR data were in good4 1 13

agreement with the data found in the literature. 7

(-)-N,N'-Bis[3-cis-endo-borneol]-N,N'-dimethyl-1,n-alkane (4.9)General procedure for the synthesis of compounds 4.9a and 4.9b . A mixture of cis-endo-N-monomethyl-3-aminoborneol ((+)-MAB) ( 4.8) (1.00 g, 5.46 mmol), 1, m-dibromoalkane ( 2.78 mmol) and K CO (0.75 g, 5.46 mmol) in 50 ml of acetonitrile was2 3

stirred and re fluxed for 16 h. The mixture was poured into 100 ml of H O and extracted2


15

with diethyl ether (3 x 50 ml). The combined organic layers were washed with brine (50ml) and dried ( Na SO ). Filtration and evaporation of the solvent gave crude 4.9 , which2 4

was purified by column chromatography (SiO , CH Cl /methanol (10:1)) resulting in2 2 2

colourless oils which solidified upon standing. The physical data for compounds 4.9aand 4.9b are as follows:(-)-N,N'-Bis[3-cis-endo-borneol]-N,N'-dimethyl-1,2-ethane (4.9a )Yield 69%. H NMR * 0.84 (s, 6H), 0.87 (s, 12H), 1.08-1.22 (m, 2H), 1.35-1.75 (m, 4H),1

1.87-2.1 2 (m, 4H), 1.9-2.4 (broad signal 4H), 2.17 (s, 6H) 2.74-2.81 (m, 2H), 2.93-2.98(m, 2H), 3.79 (dd, J = 8.9 Hz, J = 1.3 Hz, 2H). C NMR * 13.75 (q), 18.32 (q), 18.57 (t)13

19.59 (q), 25.65 (t), 41.40 (q) 48.18 (s), 48.56 (d) 50.24 (s), 55.00 (t), 65.02 (d), 74.00(d). HRMS calcd for C H N O : 392.340 found 392.340.24 44 2 2

(-)-N,N'-Bis[3-cis-endo-borneol]-N,N'-dimethyl-1,3-propane (4.9b )Yield 85%. [ "] +53.8 E (c 0.4, CH Cl ). H NMR * 0.86 (s, 6H), 0.90 (s, 12H), 1.12-1.2920 1

D 2 2

(m, 2H), 1.45-1.58 (m, 4H), 1.70-1.76 (m, 2H), 1.79-1.95 (m, 4H), 2.2-2.3 (bs, 2 x 3H,2H), 2.45 -2.55 (broad signal, 2H) 2.57 (dd, J = 8.7 Hz, J = 4.0 Hz, 2H), 3.66 (d, J = 8.7Hz, 2H). C NMR * 14.50 (q), 18.88 (q), 18.95 (t) 20.06 (q), 25.13 (t), 26.44 (t), 40.6213

(q) 45.29 (s), 48.20 (d) 50.20 (s), 54.65 (t), 65.31 (d), 73.94 (d). HRMS calcd fo rC H N O : 406.356, found 406.356.25 46 2 2

(+)-cis-endo-N-[(1-methylpyrrol-2-yl)methyl]-N-methyl-3-aminoborneol (4.11 )Secondary amino alcohol 4.10 (0.80 g, 3.07 mmol), MeI (1.5 ml, excess), NaOH (4.0 g,excess) and 1 ml of H O were successively added to 25 ml of diethyl ether. The mixture2

was stirred for 16 h and an additional 1 ml of MeI was added. After another 24 h themixture was poured into 50 ml of H O. The two layers were separated and the water2

phase was extracted with diethyl ether (3 x 30 ml). The combined organic layers werewashed with brine (50 ml) and dried (Na SO ). Filtration and evaporation of the solvent2 4

gave 0.73 g of a crude yellow solid. This material was recrystallised from 15 ml o fhexane / ethyl acetate (10:1) yielding pure 4.11 as a white solid (0.70 g, 2.55 mmol,83%); mp 54.9-56.4 EC. ["] +61.4 E (c 2.01, CH Cl ). H NMR * 0.91 (s, 3H), 0.92 (s,2 0 1

D 2 2

3H), 0.93 (s, 3H), 1.16-1.29 (m, 1H), 1.49-1.59 (m, 2H), 1.78-1.94 (m, 2H), 2.09 (s, 3H),2.64 (dd, J = 8.6 Hz, J = 3.3 Hz, 1H), 3.33 (d, J = 13.6 Hz, AB system, 1H), 3.56 (d, J =13.6 Hz, AB system, 1H), 3.66 (s, 3H), 3.75 (d, J = 8.6 Hz, 1H), 6.05 (m, 2H), 6.59 (m,1H) C NMR * 14.23 (q), 18.73 (q), 18.98 (t) 19.82 (q), 26.18 (t), 34.11 (q), 40.79 (q),13

45.25 (s), 48.12 (d) 50.12 (s), 52.45 (t), 65.48 (d), 73.86 (d), 106.39 (d), 109.94 (d) ,122.47 (d), 128.93 (s). HRMS calcd for C H N O: 276.220, found 276.220.17 28 2

(+)-cis-endo-N-[2-pyridylmethyl]-N-methyl-3-aminoborneol (4.12 )Secondary amino alcohol 4.8 (1.00 g, 5.50 mmol), picolyl chloride hydrochloride (0.90

Chapter 4

16

g, 5.50 mmol) and K CO (1.66 g, 12.0 mmol) in 50 ml of acetonitrile was stirred and2 3

refluxed for 4 days. The reaction mixture was poured into 50 ml of H O and extracted2

with diethyl ether (2 x 50 ml). The combined organic layers were washed with brine (50ml) and dried (Na SO ). Filtration and evaporation of the solvent gave 0.73 g of a crude2 4

yellow oil. C olumn chromatography (SiO , CH Cl /methanol (5:1)) afforded compound2 2 2

4.12 as a light yellow oil, which solidified upon standing. Yield 73%; mp 90.2-91.5 EC.["] +66.1 E (c 2.24, CH Cl ). H NMR * 0.88 (s, 3H), 0.89 (s, 3H), 0.92 (s, 3H), 1.12-2 0 1

D 2 2

1.29 (m, 1H), 1.48-1.63 (m, 2H), 1.78 (t, J = 4.0, 1H) 1.82-1.96 (m, 1H), 2.12 (s, 3H),2.82 (dd, J = 8.8 Hz, J = 4.0 Hz, 1H), 3.43 (d, J = 13.8 Hz, AB system, 1H), 3.74 (d, J =8.8 Hz, 1H), 3.76 (d, J = 13.8 Hz, AB system, 1H), 7.11-7.17 (m, 1H), 7.39 (d, J = 7.8 Hz,1H), 7.62 (dt, J = 7.8 Hz, J = 1.9 Hz, 1H), 8.53 (m, 1H) C NMR * 14.23 (q), 18.65 (t)13

19.81 (q), 26.18 (t), 41.15 (q), 45.05 (s), 48.11 (d) 50.22 (s), 62.69 (t), 64.54 (d), 73.93(d), 121.89 (d), 122.40 (d), 136.36 (d), 149.07 (d) 159.10 (s). HRMS calcd fo rC H N O: 274.205, found 274.204.17 26 2

Conjugate addition of diethylzinc to chalcone (4.13) or cyclohexenone (4.17) usingcatalytic amounts of Ni(acac) and chiral amino alcohols 2

This p rocedure is typical for all conjugate addition reactions described in Section 4.3.A solution of Ni (acac) (0.07 mmol) and chiral ligand (amounts, see Tables 4.1 and 4.2)2

in 2 ml of acetonitrile was stirred and refluxed for 1 h under nitrogen. In general thisresults in a clear green solution. Substrate was added (1.0-2.0 mmol), the mixture wascooled to -30 EC and diethylzinc in hexane (1 M) or toluene (1.1 M) (1.5 equivalent)was added. Stir ring was continued at -25 EC for 16 h. An aliquot of the solution (0.1 ml)was taken and quenched with 1 ml of aqueous 1 N HCl. After extraction with 1 ml ofdiethyl ether the conversion was determined by GC analysis. Retention times (ove ntempera ture 225 EC, flow 101 ml/min He): 1,3-diphenyl-2-propenone ( 4.13 ), 5.66 min;1,3-diphenylpen tan-1-one ( 4.14 ), 4.93 min; (oven temperature 100 EC, flow 101 ml/minHe): cyclo -2-hexen-1-one ( 4.17 ), 2.87 min; 3-ethylcyclohexan-1-one ( 4.18 ), 5.88 min.If com plete conversion was achieved, the mixture was poured into 25 ml of aqueous 1N HCl and extracted with diethyl ether (3 x 20 ml). The combined organic layers werewashed with brine (25 ml), dried (MgSO ), filtered and evaporated to give the crude4

1,4-products. (Caution: compound 4.18 is volatile and long evaporation times shouldbe avoided. ) After purification by column chromatography (SiO , hexane:diethyl ether2

5:1) the e.e.'s were determined. 1,3-Diphenylpentan-1-one ( 4.14 ): HPLC analysis (seeSection 3.8); 3-ethylcyclohexan-1-one ( 4.18 ): derivatisation with optically pure 1,2-diphenylethylene diamine followed by C NMR analysis, see also Chapter 5. All H13 11 1

NMR and C NMR data of 4.14 and 4.18 were in good agreement with the data found13

in Chapter 3.


17

1. Kitamura, M.; Okada, S.; Suga, S.; Noyori, R. J. Am. Chem. Soc. 1989 , 111, 4028.

2. Sacconi, L.; Mani, F.; Bencini, A. In Comprehensive Coordination Chemistry; Wilkinson, G.; Gillard,R.D.; McLeverty, J.A., Eds.; Pergamon: Oxford, 1987 , Vol. 5, Chapter 50.

3. March, J. Advanced Organic Chemistry 3th ed.; Wiley-Interscience: New York, 1985 , Chapter 10.

4. Borch, R.F.; Bernstein, M.D.; Durst, H.D. J. Am. Chem. Soc. 1971 , 93, 2897.

5. An adaptation of the procedure of Borch et al. was followed: Zydowski, T.M.; Dellaria, Jr., J.F. ;Nellans, H.N. J. Org. Chem. 1988 , 53, 5607.

6. Brewster, J.H. in Comprehensive Organic Synthesis; Trost, B.M.; Fleming,I., Eds.; Pergamon: Oxford,1991 , Vol. 8, Chapter 1.9 and references therein.

7. Tanaka, K.; Ushio, H.; Kawabata, Y.; Suzuki, H. J. Chem. Soc. Perkin Trans. 1 1991 , 1445.

8. Tanaka, K.; Matsui, J.; Suzuki, H. J. Chem. Soc., Perkin Trans. I, 1993 , 153.

9. Bolm, C.; Ewald, M. Felder, M. Chem. Ber. 1992 , 125, 1205.

10. Kindly provided by Roelof Stroetinga, research student in the group of Prof. Kellogg.

11. A. Alexakis, J. C. Frutos, P. Mangeney, Tetrahedron: Asymmetry 1993 , 4, 2431.

Acknowledgements

R. Imbos is gratefully acknowledged for the pleasant and fruitful cooperation i nconnection with her undergraduate research. A. Arnold is thanked for the performanceof some experiments described in this Chapter and R. Stroetinga for the synthesis ofcompoun d 4.16 . Mr. M. Suijkerbuijk is acknowledged for assistance with the e.e .determinat ions of 4.14 and Mr. W. Kruizinga and Dr. J. Herrema are acknowledged forthe e.e. determinations of 4.18 .


1

Chapter 5

Towards a New Catalytic System for Enantioselective Conjugate Addition of Organometallic Reagents

5.1 Introduction

As was s hown in the preceding Chapters all successful reports on catalyti cenantios elective conjugate addition of organometallic reagents are limited to on especi fic type of substrate. In spite of the introduction of novel tri- and tetradentat eamino alcohols, the nickel catalysed alkyl transfer from diethylzinc is onl yenantioselective for acyclic enones so far (Chapter 4). In this Chapter attempts will be described to develop a catalyst which is effective forcyclic and acyclic enones. Three main variations have been investigated: - The reactivi ty of diethylzinc towards enones in the presence of metal salts other

than Ni(acac) .2

- Tuning of the copper catalysed alkyl transfer from diethylzinc to enones withchiral ligands.

- Tuning of the copper catalysed methyl transfer to enones wit htrimethylaluminium as organometallic reagent.

5.2 The reactivity of diethylzinc towards enones in the presence of metalsalts

Organozinc compounds (R Zn and RZnY, I) react extremely sluggish with carbony l2

compounds. The high covalent character of the carbon-zinc bond and the relativel ymoder ate Lewis acidity of Zn(II) are responsible for this inertness. The reactivity can1

be enhanced by th e use of (chiral) ligands and/or by transmetalation to a second metal.2

The empty low-lying 4p orbitals of zinc allow many transmetalation reactions wit hmetallic salts to proceed as long as they are thermodynamically favoured. This ability1

permits the conversion of organozinc reagents into more reactive organometalli creagent s RML (II) (Eq. 5.1). The synthetic utility of this approach has bee nn

1b

demonstrated with nickel, copper, palladium and titanium salts.1, 3 1, 4 1, 5 1, 6

Y = R, halideM = Ti, Pd, Ni, CuX = halide

YX

MLnR

Zn (5.1)

II

X Zn Y+R MLn

I

R Zn Y + X MLn

Chapter 5

2

With this knowled ge we have examined several metal salts as catalysts in the conjugateadditi on of diethylzinc to chalcone (Table 5.1, entries 1-7). As chiral ligand amin oalco hol (-)-DAIB ( 5.3 , Figure 5.1), successfully applied in the nickel catalyse dconjugate ad dition, was employed. The catalytic conjugate addition reactions were runon a 1 mmol scale. In general, a solution of 7 mol% of metal salt and 16 mol% of (-)-DAIB in 2 ml of acetonitrile was heated at reflux for 1 h. Except for CuBr, all in situprepared complexes gave clear solutions, indicating the homogeneous nature of thecatalytic system. The three palladium salts gave a dark green solution and CuBr, ZnI ,2Fe(a cac) , and Co(acac) afforded orange, colourless, red, and purple solutions ,3 2

respecti vely. Chalcone ( 5.1) was added at room temperature followed by 1.5 ml o fdiethyl zinc in hexane (1 M) at -30 EC. After 16 h at -30 EC the conversion wa sdetermined by GC analysis. The results of these reactions are shown in Table 5.1.

PhPh

O

5.1 5.2

CH3CN, hexane-30°C

metal salt (cat.)chiral ligand (cat.)

Ph Ph

O

*+ Et2Zn

N

OH

N

NHO

NOH

N OH

PhPh

5.3 5.4 5.5 5.6


3

Table 5.1 Enantioselective conjugate addition of diethylzinc to chalcone (5.1)using various metal complexes.a

entry metal salt chiral ligand conv. (%) e.e. (%) abs. conf.b c d

1 Pd(OAc) 5.3 < 52

2 Pd(CH CN) Cl 5.3 < 53 2 2

3 PdCl 5.3 < 52

4 CuBr 5.3 60 41 R 5 ZnI 5.3 < 52

6 Fe(acac) 5.3 < 53

7 Co(acac) 5.3 70 67 R2

8 Co(acac) 5.4 80 83 S2

9 Co(acac) 5.5 48 33 R2

10 Co(acac) 5.6 58 28 S2

a. Reactions at - 30EC in 2 ml of acetonitrile and 1.5 ml of hexane using an in situ prepared catalyst from 7 mol%of metal salt and 16 mol% of chiral ligand (Figure 5.1). Reaction time 16 h. b. Conversion to the 1,4-product,determined by GC analysis. c. Determined by HPLC analysis: Daicel, Chiralcel OD; 0.25% iPrOH in hexane,flow rate 1.0 ml/min, UV detector (254 nm). d. Comparison of retention times of ( R)- and ( S)-5.2 with knowndata gave the absolute configuration. 7

Figure 5.1 Chiral ligands used in the Co(acac) catalysed addition of diethylzinc to2

5.1.

Chapter 5

4

In most cases the conjugate addition reaction appears not to proceed. (In all entries no1,2-addition was observed either.) Only with CuBr and Co(acac) conversions to the2

1,4-pro duct higher than 50% were achieved. In comparison to the nickel catalyse dreaction (> 95% conversion to the 1,4-product after 2 h at -30 EC) the copper and cobaltcatalysed conjugate a dditions are slow and a considerable amount (ca. 5%) of a reducedbyproduct (1,3-dip henylpropan-1-one) has been detected. In spite of the relatively low8

conversi on, product 5.2 was isolated and the e.e.'s were determined by HPLC analysis.With the CuBr and Co(acac) catalysed addition enantioselectivities were achieved of2

41% and 67%, respectively. These e.e. values indicate that (-)-DAIB is capable o fcreating selective catalysts with other metal salts, as well.A number of other chiral amino alcohols were examined with Co(acac) in the model2

reaction (Figure 5.1 and Table 5.1, entries 8-10). Employing (+)-DAB ( 5.4) as chiralligand, again the same enantioselectivity (83% e.e.) as in the nickel catalysed versionwas found for product 5.2 . Amino alcohols (+)-diphenyl(1-methylazetidin-2-yl) -methanol ( 5.5) and (2 S,2'S)-2-(hydroxymethyl)-1-[(1-methylpyrrolidin-2yl)methyl]-9

pyrrolidine ( 5.6) gave substantial lower conversions to the 1,4-product, moderate e.e.'s,and the reduced byproduct was detected in higher amounts (ca. 20%).Although reasonable y ields and e.e.'s were found in the Co(acac) / chiral amino alcohol2

catalyse d additions, no further experiments were performed for the following reasons:- The addition shows lower regioselectivity for the 1,4-product compared to the

Ni(acac) / chiral amino alcohol catalysed reaction.2

- With cyclohexenone (5.7) as substrate and (-)-DAIB as chiral ligand the reactionproceeds sluggish and no enantioselectivity was found.

- Co pper salts showed to be more promising in an attempt to achieve catalyti cenantioselective conjugate additions to cyclic and acyclic enones ( vide infra).

5.3 Copper catalysed enantioselective addition of diethylzinc to enones

IntroductionAs was shown in the preceding Section the catalyst derived in situ from CuBr and (-)-DAIB is capable of catalysing the conjugate addition of diethylzinc to chalcone. Theregioselective co nversion to the 1,4-product is slow (compared to Ni(acac) / (-)-DAIB)2

and a moderate enantioselectivity was found (41% e.e., Table 5.1, entry 4). Althoughthis re sult supposes quite the contrary, copper salts seem to be the metal salt of choicefor achieving catalytic enantioselective conjugate addition to both cyclic and cyclicenones for the following reasons: (1) Probably due to the affinity of the (chiral) nickelcenter for the carbonyl oxygen, only enantioselective alkyl transfer occurs in the case

VIVIII

ONiL*n

R R

ONiL*n

CuL*nR''

O

R R'


5

of s-cis enones, i.e. chalcone, as is shown in Figure 5.2 ( III ). For cyclohexenone ,probably intermediate IV will be formed with the chiral nickel complex too far awayfrom the $-position and therefore not able to introduce any asymmetry (see precedingChapters). Copper complexes, on the other hand probably coordinate to the carbon-carbon double bond of the enone, furnishing intermediates like V, with possibilities10

for enanti oselective alkyl transfer to both cyclic and acyclic enones ( i.e. s-trans and s-cis enones).

Figure 5.2 Possible intermediates in the nickel catalysed alkyl transfer to acyclic(III) and cyclic enone (IV) and the copper catalysed reaction (V).

(2) P reliminary investigations on the CuI catalysed addition of diethylzinc t ocyclohexenone ( 5.7) have revealed that this reaction is slow and not regioselective. SeeGC diagram A in Figure 5.3, which shows besides the 1,4-product ( 5.8 , retention timeof 5.88 min), several other products, and a substantial amount of unreacted substrate(retention time of 2.88 min). With an additional (chiral) ligand, for example (+)-DAB(5.4), the conversion is faster and a higher regioselectivity to the 1,4-product wa sfound. (ca. 80%, GC diagram B, Figure 5.3). This is a typical example of ligand -acceler ated catalysis and "the gate to successful metal catalysed asymmetri creactions" . In reactions with transition metals, however, a variety of metal complexes11

often exists simultaneously in solution. These molecular assemblies for m12

spont aneously and the composition of the mixtures is dictated by thermodynami cfactors. The goal in the development of efficient asymmetric catalysis is to find thatparticular, active and highly enantioselective complex.(3) Furthermore, a switch to copper salts was stimulated by a report of Alexakis and co-workers, of an in situ prepared complex of CuI and a chiral trivalent phosphorus ligand,which catalyses the enantioselective conjugate addition of diethylzinc t ocyclohexeno ne ( 5.7 , Scheme 5.1). However, with chalcone as substrate no e.e. for the13

1,4-product was found.With this knowledge we have examined three different classes of chiral ligands in theCuI (10 mo l%) catalysed addition of diethylzinc to cyclohexenone ( 5.7) and t ochalcone ( 5.1).

L* =

5.7

NP

OPh

N

i-Pr5.9

70 % yield32 % e.e. (S)

O

Et

O

5.8

*+ Et2Zn CuI (cat.)L* (cat.)

toluene20°C


7

Figure 5.3 GC diagrams of the CuI catalysed (A, after 5 days at -10EC in toluene)and the CuI / (+)-DAB (5.4) catalysed (B, after 24 h at -20EC in toluene)addition of diethylzinc to cyclohexenone.

Scheme 5.1 CuI catalysed addition of diethylzinc to 5.7, reported by Alexakis and co-workers.13

Amino alcohol ligandsAs already mentione d the chiral tertiary amino alcohol (+)-DAB or (-)-DAIB (20 mol%)showed a su bstantial ligand-acceleration in the copper catalysed conjugate addition ofdiethylzinc to 5.7 (Figure 5.3). The reaction time was shortened and the regioselectivityto the 1,4-produc t has been improved to ca. 80%. Unfortunately, with both chiral aminoalcohol ligands, the 1,4-product 5.8 was isolated without any enantioselectivity, i ncontrast to the copp er catalysed addition to chalcone (Table 5.1, entry 4). The e.e. of theresulti ng 1,4-product was determined by the formation of diastereomeric aminals withcommer cially available optically pure 1,2-diphenylethylene diamine. This method is14

mu ch faster than the routinely accomplished formation of diastereomeric ketal saccording to the Hiemstra-Wynberg method. 15

Phosphorus ligandsTraditionally phosphorus compounds are considered one of the best soft base ligandsfor copper(I) and a significant number of complexes of known crystal structure ,primarily with trivalent monodentate ligands, has been described. In general, th e16

monomeric phosphorus ligands only coordinate as single ligand donors and are no tin volved in any bridging role. Furthermore, trivalent phosphorus compounds ar eknown as (chiral) lig ands for stoichiometric conjugate organocopper additions to cyclicenones. 17 ,18 ,19

benzeneRT, 3 h

HMPT+OO

P NOHOH

(5.2)

5.10

Chapter 5

8

Recentl y, Hulst has synthesised quite easily from ( S)-2,2'-binaphthol and HMPT, anovel chiral trivalen t phosphorus compound ( 5.10 , Eq. 5.2). Phosphorus amidites have20

been app lied extensively as improved capping reagents for the synthesis o foligonuc leotides, however, they represent a class of compounds, hardly recognized21 22

as ligands for catalytic transformations. 23

When ligand 5.10 (20 mol%) was examined in the CuI (10 mol%) catalysed addition ofdi ethylzinc to cyclohexenone ( 5.7 , Scheme 5.1) the above mentioned ligand -acceleration is even more pronounced. Within 24 h the catalytic reaction, in toluene /hexane at -10 EC, results in selective formation of the 1,4-product (> 90%, GC analysis).Only traces of reduced products were detected. Product 5.8 was isolated in 75% yieldwith an e.e. of 35% (Table 5.2, entry 1).


9

Table 5.2 . Enantioselective conjugate addition of diethylzinc to cyclohexenone andchalcones catalysed by ligand 5.10 and copper(I) salts.a

entry substrate CuX (mol%) mol% of 5.10 solvent e.e. (%) b c

abs. conf. d

1 5.7 CuI (10) 20 toluene / hexane 35 S 2 5.7 CuI (10) 20 THF / hexane 22 S 3 5.7 CuI (10) 20 diethylether / hexane 0

4 5.7 CuI (10) 20 acetonitrile / hexane 5S

5 5.7 CuBr (10) 20 toluene / hexane 35 S 6 5.1 CuI (10) 20 toluene / hexane 14 R 7 5.1 CuI (10) 20 toluene 47 R 8 5.7 CuI (10) 20 toluene 38 S 9 5.7 CuI (10) 50 toluene 18 Se

10 5.7 CuI (10) 10 toluene 25 S11 5.7 CuI (10) 15 toluene 35 S12 5.7 CuI (50) 100 toluene 40 Sf

a. Reactions at 1 mmol scale at -10 EC. Reaction time 24 h. Conversion to the 1,4-product ( 5.8) > 90% (basedon GC analysis). b. 1.5 Equivalent of diethylzinc added as solution in hexane (1M) or toluene (1.1M) c .Enantiomeric excess of 3-ethylcyclohexanone ( 5.8) determined by derivatisation with optically pure 1,2 -diphenylethylene diamine. E.e. determination of 5.2, see Table 5.1. d. Comparison of the optical rotation of14

5.8 with kno wn data gave the absolute configuration. For e.e. determination of 5.2, see Table 5.1. e .24

Conversion to the 1,4-product (95%) only achieved after an additional 3 days at RT. f. At 0.5 mmol scale in15 ml of toluene.

The enantioselectivity in the formation of the 1,4-product showed to be strongl ydependent on the solvent used. Although the reaction is also very selective to the 1,4-product in THF, diethyl ether, and acetonitrile significant lower (or no) e.e.'s wer efound (22, 0, and 5%, respectively, entries 2-4). With CuBr instead of CuI the sam ereaction rate and enantioselectivity was found.When the most selective reaction conditions, found for 5.7 , were employed with theacycli c substrate chalcone ( 5.1) a very regioselective 1,4-addition occurred .Unfortunately, for 5.2 , isolated in 87% yield, a disappointing e.e. of 14% was found(entry 6). At this stage all conjugate addition reactions described in this Chapter wereperformed with a solution of diethylzinc in hexane. When a solution of diethylzinc in

Chapter 5

10

tolu ene was added to the reaction mixture a remarkable increase in e.e. for 5.2 wasfound (entry 7). With cyclohexenone this variation had a minor influence (entry 8). Apossible explanation is B-stacking of the chiral catalyst and the aromatic substrat epromoted by toluene into a more enantioselective aggregate.Employing 50 mol% of 5.10 and 10 mol% of CuI, furnished a chiral catalyst which isonly activ e at room temperature, less regioselective and hardly enantioselective (entry9). Better results were obtained with 5.10 / CuI ratios of 1 and 1.5 (entries 10 and 11).At higher c oncentration of the chiral catalyst no significant enhancement was found inreactivity and enantioselectivity (entry 12). These remarkable results indicate the highactivity of t he enantioselective catalyst, prepared in situ of appropriate amounts of CuIand 5.10 , compared to the other possible complexes in solution.Although moderate enantioselectivities were found for 5.2 and 5.8 these preliminaryexperiments represe nt the first example of catalytic enantioselective conjugate additionof an organo metallic reagent to both a cyclic and acyclic substrate and provide severalapproaches for further investigation (see Chapter 6).

Sulfur ligandsConsistent with the soft base behaviour of sulfur ligands, a considerable number o fcopper(I) complexes of sulfur ligands has been reported. In general, it is a goo dmonodenta te ligand in trigonal planar and tetrahedral complexes, but it is equally wellfunctioning as a bridging ligand in dinuclear and polynuclear aggregates. Sulfu r16

containing compounds are known as chiral ligands for stoichiometric and catalytic25

en antioselective conjugate addition reactions of Grignard reagents to enones. 26

Furthermore, the research group of Prof. Kellogg has shown that optically pure thiolsand sulfides are successful catalysts in the enantioselective addition of diethylzinc tobenzaldehyde. Therefore, several (novel) sulfur containing compounds wer e27

exami ned as chiral ligands in the copper catalysed addition of diethylzinc t ocyclohexenone and chalcone.First, thiophosphonate 5.11 (see Figure 5.3), known as a chiral equivalent of H S,28 29

2

was tested in the CuI catalysed addition of diethylzinc to cyclohexenone. A mixture ofCuI (10 mol%) and of 5.11 (15 mol%) in toluene furnished a catalyst which is ratherslow and less regioselective to the 1,4-product (compared to 5.10 ). After 4 days at -10EC, 88% of the substrate was converted to 5.8 with an e.e. of 15%. Probably due tosolu bility problems this is not the chiral ligand of choice under these reactio nconditions.

5.13

S N

O

5.11

OO

PS

SH

5.12

N

S

OPh

R1N

R2

a R1 = Me, R2 = Hb R1 = Me, R2 = Mec R1 = i-Pr, R2 = H

N

S

OPh

N

5.12d

*

5.8

O O

5.7

+ Et2ZnCuOTf (cat.)5.12 (cat.)

toluene-10°C


11

Figure 5.3 Sulfur containing ligands used in the copper catalysed addition ofdiethylzinc to cyclohexenone and chalcone.

Later on in this project, solubility of the in situ prepared chiral catalysts was enhancedwith CuOTf or Cu(OTf) as the copper source (see Chapter 6). When a mixture o f2

Cu(OTf) (3 mol%) and of 5.11 (6.5 mol%) in toluene was employed as chiral catalyst,2

the addition of diethylzinc to cyclohexenone proceeds within 16 h at -15 EC and withgood selectively to the 1,4-product. Product 5.8 was isolated in 65% yield, however,no significant e.e. (< 10%) was found. For chalcone ( 5.1) this Cu(OTf) / 5.11 catalyst2

was less reactive (55% isolated yield after 2 days at -15 EC) and a disappointing e.e of7% was found.

Table 5.3 Enantioselective 1,4-addition of diethylzinc to cyclohexenonecatalysed by CuOTf and sulfur containing ligands 5.12. a

entry ligand e.e. of 5.8 (%) abs. conf.b b

1 5.12a 47 R 2 5.12b 39 R 3 5.12c 62 R 4 5.12d 49 R

Chapter 5

12

a. Reactions at 1 mmol scale at -10 EC in 5 ml of toluene. Reaction time 2-14 h (see text). Conversion to the 1,4-product > 95% (based on GC analysis). Isolated yields > 70%. b. Determination of e.e. and absolut econfiguration of 3-ethylcyclohexanone ( 5.8), see Table 5.2.

Next, several novel substituted thiazolidin-4-ones 5.12a-d (Figure 5.3), synthesised byHof, were examined in the CuOTf catalysed addition of diethylzinc to cyclohexenone30

(Table 5.3). The synthesis of 5.12a-d is based on cyclocondensation of an "-mercaptoacid, ani line, and the corresponding carboxaldehyde furnishing the thiazolidin-4-onesin qua ntitative yield as a mixture of diastereoisomers. Diastereomerically pure cis-5.12a-d were obtained by recrystallisation. With ligand 5.12a (11 mol%) and 5 mol%30

of CuOTf a very selective reaction to the 1,4-product (> 90%, GC analysis) was found.Isolation, purification, and derivatisation of 5.8 revealed an e.e. of 47% (entry 1).In order to improve the observed enantioselectivity sterically demanding groups onboth the pyridyl group ( 5.12b and 5.12d ) and the "-position ( 5.12c ) were introduced.Clearly, a methyl substituent on the pyridine ring reduces the e.e. ( 5.12b , 39%, entry2), probably by inducing the formation of another aggregate of the copper complex inso lution. Quinoline, instead of pyridine ( 5.12d ), had a minor influence, whereas asterically deman ding group at the "-position in ligand 5.12c gave the best results. After2 h quantitative conv ersion to the 1,4-product was achieved and an enantiomeric excessof 62% was observed.For comparison, the chiral oxazoline substituted thiophene 5.13 was also examined31

in the model reaction given in Table 5.3. Again a high selectivity of 95% to the 1,4-product was observed. However, the low enantiomeric excess of 5.8 (5%) found withthis chi ral thiophene, indicates the necessity of the chiral cavity created by the cisconfiguration of both substituents in ligands 5.12 . Furthermore, the role of the amidefunctionality in ligands 5.12 is not clear at this moment.With chalcon e as substrate complete 1,4-selectivity was observed as well in the CuOTf/ 5.12 catalysed addition, however, low e.e.'s for 5.2 (up to 11% for ligand 5.12c ) wereobserved. In cooperation with the group of Prof. Kellogg, optimisation of the structureof this novel type of sulfur containing ligands is under progress. The scope of thi sreaction and possible other asymmetric transformations will be explored.

5.4 Enantioselective copper catalysed methyl transfer to enones withtrimethylaluminium as organometallic reagent

IntroductionThe 1 ,4-addition of hydrocarbon substituents to ",$-unsaturated carbonyl compoundsis usually achieved by using organocuprate reagents. Asymmetric conjugate addition32

Ph Ph

O

*

5.25.1

PhPh

O

+ EtMgBr


13

of organocupr ates is well established with chiral and prochiral substrates. In an effort33

to create a system capable of catalytic enantioselective conjugate addition of Grignardreagents, preliminary experiments were performed in the presence of (-)-DAIB ( 5.3).Contrary to other substrates, the addition of EtMgBr to chalcone in diethyl ether at34

room temperature proceeded quite selective to the 1,4-product. (GC analysis revelaeda regioselec tivity of ca. 90%, Table 5.4, entry 1). In the presence of catalytic amounts35

of 5.3 and KO t-Bu or Ni(acac) in diethyl ether / propionitrile at -50 EC, this 1,4-addition2

is even more regioselective (> 90%, GC analysis) and 5.2 could be isolated in 86% and92%, respectively (entries 2 and 3). Unfortunately, in both cases no enantioselectivealkyl transfer occurred.Since there i s no enantioselectivity at all in the presence of (-)-DAIB and secondly, theena ntioselective transfer of alkyl groups from Grignard reagents to enones is stil llimited to one type of substrate (see Section 2.3), the main point of our research hasbeen shifted to other organometallic reagents.

Table 5.4 Conjugate addition of EtMgBr to chalcone.a

entry chiral ligand (mol%) metal source (mol%) yield (%) e.e.b

(%) c

1 - - nd -c

2 5.3 (16) KO t-Bu (16) 86 0 3 5.3 (16) Ni(acac) (7) 92 22

a. Reaction conditions see text. b. Isolated yield. c. Yield not determined. GC analysis revealed a 1,4 -selectivity of 90% at 100% conversion.

Trimethylaluminium as organometallic reagentThe 1 ,4-addition of hydrocarbon substituents to ",$-unsaturated carbonyl compoundshas a lso been reported with organomanganese reagents, organotitanates,36 37

organoaluminium halides and trialkylaluminium reagents. Especially the highly38 39 ,40

se lective copper(I) catalysed alkyl transfer from aluminium compounds to severa l

PhPh

O

Ph Ph

O

*+ Me3Al

L* (cat.)CuBr (cat.)

solvent-30°C5.1 5.2a

Chapter 5

14

cyc lic and acyclic enones, reported by the research group of Kabbara an dWest ermann, catched our attention. Screening of chiral ligands with affinity fo r40c,d

copper(I), shou ld furnish a novel catalytic enantioselective conjugate addition method.Therefore, amino al cohols 5.3, 5.4 and 5.6 , phosphorus amidite 5.10 and thioephedrines5.14 and dimer 5.15 were examined in the CuBr catalysed addition o f30a

trimethyla luminium (TMA) to chalcone. The results are summarised in Table 5.5. Withligands 5.3, 5.4 and 5.6 the addition reaction in ethyl acetate is highly regioselectiveto the th e 1,4-product, and comparable with those found in the literature withou taddition al ligand. Enantioselectivities for 5.2a of 5%, 30%, and 5% were observed,40c,d

respectiv ely. Although the e.e. found for 5.2a , when ligand 5.4 was employed, i smoderate, this is to our knowledge the first example of a catalytic enantioselective alkyltransfer from aluminium compounds to an enone. 41

Table 5.5 CuBr catalysed enantioselective conjugate addition of Me Al to3

chalcone.a

entry ligand solvent conv. (%) e.e. (%)b c

1 5.3 ethyl acetate 91 5 2 5.4 ethyl acetate 94 30 3 5.6 ethyl acetate 90 5 4 5.10 ethyl acetate 68 ndd e

5 5.14 ethyl acetate 55 < 5d

6 5.15 ethyl acetate 66 ndd e

7 5.4 THF 98 8 8 5.4 acetonitrile 96 0

a. Reactions at 1 mmol scale in 2 ml of solvent using an in situ prepared catalyst from 7 mol% of CuBr and 20mol% of chiral ligand; 0.7 ml of 2M solution of TMA in hexane (or toluene) was added. Reaction time 48 h.b. Conversion to the 1,4-product, determined by GC analysis. c. Determined by HPLC analysis, see Table 5.1.d. Reaction proceeds only at RT. Conversion after 3 days. e. E.e. could not be determined due to isolation andpurification problems.

S N

Ph

H(

2HS N

Ph

H.HCl

5.14 5.15

(5.3)

5.1

PhPh

O

+ Me3Al

L* (cat.)CuBr (cat.)

toluene Ph Ph

OH

5.16

5.185.175.8a

O

OO OH

+ +

TMA5.4 (cat.)CuBr (cat.)

O

5.7

(5.4)ethyl acetate


15

When phosphorus amidite 5.10 or thiolephedrines 5.14 and 5.15 were employed a schiral ligands, the addition of TMA to 5.1 proceeded very sluggish and no significante.e. for 5.2a was observed (entries 4-6). The transfer of the methyl group of TMA to thecopp er atom and subsequently to the enone is apparently hampered by these chira lligands. Furthermore, the copper catalysed addition of TMA to enones showed a remarkablesolvent dependency. The conjugate addition of TMA to 5.1 , in the presence of ligand5.4 an d CuBr, proceeded smoothly in other polar solvents like THF or acetonitrile .Howeve r, hardly any enantioselectivity was observed. Coordinative solvents open thedimeric s tructure of organoaluminium compounds to form monomeric complexes witha signi ficantly reduced electronegativity of aluminium. The weak nucleophilicity of40d

the organometallic reagent and as a result complete 1,4-addition can be attributed tothis property.On the o ther hand, reaction of TMA with enones in apolar solvents like toluene o rhexane results in fast 1,2-addition to the carbonyl group (Eq. 5.3; see also ref. 40d).Compound 5.16 was isolated in 90% yield. In spite of the presence of 5.4 , this carbonyladdition proceeded without any enantioselectivity.

Chapter 5

16

Employ ing the most successful conditions found in the reaction with chalcone, th ecopper ca talysed addition of TMA to cyclohexenone ( 5.7) resulted in a mixture o fproducts. Besides traces of reduced products, 5.8a , the 1,2-product 5.17 , and a dimericstructure, probably 5.18 , were detected by GCMS analysis (Eq. 5.4). Apparently, withcyclic substrates the 1,4-addition of TMA is competing with several other reaction sincluding a tandem reaction of the enolate with unreacted starting material. Since40d

there is no selectivity to a single product with 5.7 was usedas substrate no furthe rexperiments were performed using other cyclic substrates.


It has been demons trated that transmetalation of alkyl group from dialkylzinc to copperor cobalt and su bsequently to the ",$-unsaturated ketone is possible. However, with thechiral amino alcohol ligands 5.3-5.6 no new general catalysts were obtained capableof enantioselective conjugate addition to both cyclic and acyclic enones. Better catalysts were obtained with the combination of CuI and phosphorus amidit e5.10 or sulfur containing ligands 5.12 . With both ligands a remarkable ligand -acce leration was observed, resulting in very regioselective conjugate additions o fdiethylzinc to cyclohexenone and chalcone. The system, derived of CuI and tw oequiv alents of 5.10 , is the first catalyst reported so far, capable of enantioselectiv econjuga te addition to both cyclic and acyclic enones. Also, with the substitute dthi azolidin-4-ones 5.12 relatively high enantioselectivities were observed for th econjugate addition of diethylzinc to cyclohexenone.Int eresting results were achieved with the CuBr catalysed enantioselective methy ltransfer from trimethylaluminium (TMA) to chalcone. With chiral amino alcohol (+)-DAB (5.4) the highly regioselective 1,4-addition of TMA to chalcone proceeded with30% e.e. Unfortunately, with cyclohexenone no regioselective addition occurred. Especially the preliminary experiments on the copper / phosphorus amidite catalysedasymmetric 1,4-addition of diethylzinc to cyclohexenone and chalcone, although e.e.'sare moderate, provide several approaches for further investigation.


For general remarks, see Section 3.8.

Materials


17

The follow ing compounds were commercially available and used without purification:All metal salts sho wn in Table 5.1 (Aldrich), 1,2-diphenylethylene diamine (Fluka), andtrim ethylaluminium (2 M in hexane or toluene, Aldrich). Hexamethylphosphoru striamide (HMPT) was purchased from Aldrich and distilled before use.O,O'-(1,1'-Dinaphthyl-2,2'-diyl)dithiophosphoric acid ( 5.11 ) was prepared accordingto a published procedure. Chiral ligands (+)-diphenyl(1-methylazetidin-2-yl) -28

methanol (5.5), substituted thiazolidin-4-ones 5.12 , (4S)-4- i-propyl-2-(2-thienyl) -9 30

1,3-oxazoline (5.13 ) and thiolephedrines 5.14 and 5.15 were kindly provided by31 30a

Prof. Martens, R. Hof, C. Zondervan, and R. Hof, respectively. For all other materials,see Section 3.8.

Conjugate addition of diethylzinc to chalcone (5.1) using catalytic amounts of metalsalts and chiral amino alcoholsThis procedure is typical for all entries in Table 5.1. A solution of 0.07 mmol of metalsalt and 0.16 mmol of chiral amino alcohol in 2 ml of acetonitrile was stirred an drefluxe d for 1 h under nitrogen. The solution was cooled to room temperature and 208mg (1.0 mmol) of 1,3-diphenyl-2-propen-1-one ( 5.1) was added. The mixture wa scool ed to -35 EC and 1.5 ml of diethylzinc in hexane (1 M, 1.5 mmol) was added .Stirring was continued at -30 EC for 16 h. An aliquot of the solution (0.1 ml) was drawnand quenched with 1 ml of 3 N HCl. After extraction with 1 ml of diethyl ether th econversion was determined by GC analysis. Retention times (oven temperature 225 EC,flow 101 ml/min He) : 1,3-diphenyl-2-propenone ( 5.1), 5.66 min; 1,3-diphenylpentan-1-one (5.2), 4.93 min. If the conversion to 5.2 was higher than 50%, the mixture wa spoured into 25 ml of aqueous 3 N HCl and extracted with diethyl ether (3 x 20 ml). Forpurifica tion and e.e. determination of 5.2 , see Section 3.8. Conversions and e.e. valuesare given in Table 5.1.

O,O'-(1,1'-Dinaphthyl-2,2'-diyl)-N,N-dimethylphosphorus amidite (5.10)The synthesis of 5.10 is somewhat improved compared to the published procedure. 20

To a mixture of ( S)-(-)-1,1'-bi-2-naphthol (1.81 g, 6.3 mmol) in 10 ml of dry benzenehexa methylphosphorus triamide (1.14 g, 7.0 mmol) was added. After 1 min a whit esolid prec ipitated. The mixture was stirred for 3 h at ambient temperature and the solidwas collected by f iltration, washed with diethyl ether (10 ml) and dried in vacuo to give1.99 g (5. 5 mmol, 88%) of pure 5.10 . An optical rotation of [ "] = +593 E (c 0.90 ,2 0

D

CHCl ) was fou nd contrary to an earlier report. All other spectroscopic data ( H NMR,320 1

C NMR, P NMR, HRMS) were in good agreement with reported values.13 31 20

Conjugate addition of diethylzinc to cyclohexenone (5.7) or chalcone (5.1) using

Chapter 5

18

catalytic amounts of CuI and chiral amino alcohols, phosphorus amidite 5.10 orsulfur containing ligands 5.11-13This p rocedure is typical for all conjugate addition reactions described in Section 5.3.A solution of copper salt and of chiral ligand in 5 ml of toluene was stirred at ambienttemperature for 1 h under nitrogen (for amounts of chiral catalyst, see text in Section5.3 and Tables 5.2 and 5.3). In general this results in a clear solution. If not, a nappropriate am ount of CH Cl was added untill a clear solution was obtained. Substrate2 2

was ad ded (1.0-2.0 mmol), the mixture was cooled to -30 EC and diethylzinc in hexane(1 M) or toluene (1.1 M) (1.5 equivalent) was added. Stirring was continued a tappro priate temperature (see text) for 24 h. An aliquot of the solution (0.1 ml) wa staken and quenche d with 1 ml of aqueous 1 N HCl. After extraction with 1 ml of diethylether the conversion was determined by GC analysis. Retention times (ove ntemperature 100EC, flow 101 ml/min He): cyclo-2-hexen-1-one ( 5.7), 2.87 min; 3 -ethylcyclohexan-1-one (5.8), 5.88 min. For retention times of 5.1 and 5.2 , see above.When complete conve rsion was achieved the mixture was poured into 25 ml of aqueous1 N HCl and ex tracted with diethyl ether (3 x 20 ml). If the conversion was no tcomplete, longer reaction times and / or higher temperatures were required (see text).The combined organic layers were washed with brine (25 ml), dried (MgSO ), filtered4

and evap orated to give the crude 1,4-products. (Caution: compound 5.8 is volatile andlong evaporation times should be avoided.) After purification by colum nchromat ography (SiO , hexane:diethyl ether 5:1) the e.e.'s were determined. For 3 -2

ethylcyclohexan-1-one ( 5.8): Derivatisation with optically pure 1,2-diphenylethylenediamine in CDCl (10 min, with some 4Å mol sieves) followed by C NMR analysis.3

13 14

For 1,3-di phenylpentan-1-one ( 5.2): HPLC analysis (see Section 3.8). H NMR and C1 13

NMR data of 5.2 and 5.8 were in good agreement with the data found in Chapters 3 and4. E.e. values are given in Section 5.3 and Tables 5.2 and 5.3.

Conjugate addition of EtMgBr to chalcone (5.1) using catalytic amounts of KOt-Buor Ni(acac) and (-)-DAIB (5.3)2

A solution of 0.16 mmol of KO t-Bu or 0.07 mmol of Ni(acac) and 0.16 mmol of (-)-2

DAIB (5.3) in 5 ml of propionitrile was stirred for 1 h under nitrogen. Chalcone (208mg, 1.0 mmol) was added and the mixture was cooled to -50 EC and EtMgBr in diethylether (0.4 M, 2.0 m mol) was added. After 16 h at -50 EC the reaction mixture was pouredinto 50 ml of saturated NH Cl solution, separated and extracted with diethyl ether (2 x4

25 ml). The combined organic layers were washed with brine (25 ml), dried (MgSO ),4

fil tered and evaporated to give the crude 1,4-product. For purification and e.e .determination of 5.2 , see Section 3.8.


19

Conjugate addition of trimethylaluminum (TMA) to chalcone (5.1) using catalyticamounts of CuBr and chiral amino alcohols, phosphorus amidite 5.10 orthiolephedrines 5.14 or 5.15This pro cedure is typical for all conjugate addition reactions described in Table 5.5. Asolution of 0.07 mmol of CuBr and 0.20 mmol of chiral ligand in 2 ml of solvent wassti rred at ambient temperature for 1 h under nitrogen. Chalcone was added (1.0-2. 0mmol), the mixture wa s cooled to -30 EC and TMA in hexane (2 M) or toluene (2 M) (1.5equiv.) w as added. Stirring was continued at -30 EC for 48 h. An aliquot of the solution(0.1 ml) was taken and quenched with 1 ml of aqueous 1 N HCl. After extraction with1 ml of diethyl ether the conversion was determined by GC analysis. Retention times,see above. When complete conversion was achieved the mixture was poured into 25 mlof aqueou s 1 N HCl and extracted with diethyl ether (3 x 25 ml). If conversion was notcomplete, longer reaction times and / or higher temperatures were required (see text).The combined organic layers were washed with brine (25 ml), dried (MgSO ), filtered4

and ev aporated to give the crude 1,4-products. For purification and e.e. determinationof 5.2a , see Section 3.8. Conversions and e.e. values are given in Table 5.5.

1,3-Diphenyl-1-buten-3-ol (5.16)A solution of 0.07 mmol of CuBr and 0.20 mmol of (+)-DAB ( 5.4) in 2 ml of toluenewas stirred at ambient temperature for 1 h under nitrogen. Chalcone was added (1.0mmol), the mixture was cooled to -30 EC and TMA in toluene (2 M, 1.5 equivalent) wasadded. Sti rring was continued at -10 EC for 16 h and the mixture was poured into 50 mlof aqueous 1 N HCl and extracted with diethyl ether (3 x 25 ml). The combined organiclayers were washed with brine (25 ml), dried (NaSO ), filtered and evaporated to give4

5.16 , which was purified by column chromatography (SiO , hexane:diethyl ether 3:1).2

Yield 90%. H NMR (300 MHz, CDCl ) * 1.81 (s, 3H), 2.19 (bs, 1H), 6.54 (d, J = 16.0 13

Hz, 1H), 6.69 (d, J = 16.0 Hz, 1H), 7.27-7.56 (m, 10H). C NMR * 29.82 (q), 74.68 (s),13

125.29 (d), 126.59 (d ), 127.09 (d), 127.64 (d), 127.70 (d), 128.33 (d), 128.59 (d), 136.36(d), 136.71 (s), 146.61 (s). HRMS calcd for C H O: 224.120, found 224.120.16 16

E.e. of 5.16 was determined by HPLC analysis; Daicel (Chiralcel OD), 10% iPrOH inhex ane, flow rate 1.0 ml/min, UV detector (254 nm); retention times 10.93 min an d13.54 min.

Copper catalysed addition of trimethylaluminum (TMA) to cyclohexenone (5.7) A solut ion of 0.07 mmol of CuBr and 0.20 mmol of (+)-DAB ( 5.4) in 2 ml of ethy lacetat e was stirred at ambient temperature for 1 h under nitrogen. Cyclohexenone wasadded (1.0 mmol), the mixture was cooled to -30 EC and TMA in toluene (2 M, 1. 5equivalent) was added. Stirring was continued at -30 EC for 2 days and the mixture was

Chapter 5

20

1. a) Boersma, J. In Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F.G.A.; Abel, E.W.,Eds.; Pergamon: Oxford, 1982 , Vol. 2, Chapter 16. b) Knochel, P.; Singer, R.D. Chem. Rev. 1993 , 93,2117.

2. Reviews on catalytic asymmetric diethylzinc addition to aldehydes, see: a) Noyori, R.; Kitamura, M.Angew. Chem. Int. Ed. Engl. 1991 , 30, 49. b) Soai, K.; Niwa, S. Chem. Rev. 1992 , 92, 833.

3. a) Kumada, M. Pure Appl. Chem. 1980 , 52, 669. b) Greene, A.E.; Lansard, J.P.; Luche, J.L.; Petrier, C.J. Org. Chem. 1984, 49, 931. c) Petrier, C.; de Souza Barbosa, J.C.; Dupuy, C.; Luche, J.-L. J. Org. Chem.1985 , 50, 5761. See also reference 5c.For nic kel catalysed asymmetric conjugate addition reactions of dialkylzincs, see Section 2.5 an dChapters 3 and 4.

4. a) Nakamura, E.; Aoki, S.; Sekiya, K.; Oshino, H.; Kuwajima, I. J. Am. Chem. Soc. 1987 , 109, 8056. b)Knochel, P.; Yeh, M.C.P.; Berk, S.C. Talbert, J. J. Org. Chem. 1988 , 53, 2390. c) Tamaru, Y.; Tanigawa,H.; Yamamoto, T.; Yoshida, Z. Angew. Chem., Int. Ed. Engl. 1989 , 28, 351. d) Zhu, L.; Wehmeyer, R.M.;Rieke, R.D. J. Org. Chem. 1991 , 56, 1445. e) Rozema, M.J.; AchyuthaRao, S.; Knochel, P. J. Org. Chem.1992 , 57, 1956. See also reference 13.

5. a) Neg ishi, E.; King, A.O.; Okuda, N. J. Org. Chem. 1977 , 42, 1821. b) Negishi, E.; Valente, L.F. ;Kobayashi, M. J. Am. Chem. Soc. 1980 , 102, 3298. c) Negishi, E. Acc. Chem. Res. 1982 , 15, 340.

6. a) Yoshioka, M.; Kawakita, T.; Ohno, M. Tetrahedron Lett. 1989 , 30, 1657. b) Seebach, D.; Behrendt,L.; Felix, D. Angew. Chem., Int. Ed. Engl. 1991 , 30, 1008. c) Duthaler, R.O.; Hafner, A. Chem. Rev.1992 , 92, 807. d) Rozema, M.J.; Eisenberg, C.; Lütjens, H.; Ostwald, R.; Belyk, K.; Knochel, P .Tetrahedron Lett. 1993 , 34, 3115.These publications report about chiral titanium complexes, capable of catalysing the addition o f(functional ised) dialkylzinc reagents to aldehydes. In the presence of catalytic amounts of trans-1(R),2(R)-bis(trifluoromethanesulfonamido)cyclohexane and Ti(O t-Bu) this protocol can b e6a,d

4

extended to enantioselective conjugate addition reactions to chalcone ( 5.1, see below). For th e

poured into 50 ml of aqueous 1 N HCl, separated and extracted with diethyl ether (2 x25 ml). The combined organic layers were washed with brine (25 ml), dried (NaSO ),4

filtere d and evaporated. GC and GCMS analysis revealed several products with n oselectivity to one single product.

Acknowledgement

The pleasant cooperation with Dr. R. Hof to examine sulfur containing ligands in thecopper catal ysed conjugate addition reactions is gratefully acknowledged. Dr. R. Hulstis thanked for the synthesis of 5.10 and A. Arnold for the performance of som eexperiments de scribed in this Chapter. The research group of Prof. Martens, Universityof Oldenburg and C. Zondervan are acknowledged for the synthesis of 5.5 and 5.13 ,respectively. M r. M. Suijkerbuijk and Mr. W. Kruizinga are thanked for assistance withthe many e.e. determinations of 5.2(a) and 5.8 , respectively.

5.7 Refererences and notes

43% e.e.

Ph Ph

O( )5

toluene, -15°C

(10 mol%)

N

N

Tf

Tf

Ti(Ot-Bu)2

Oct2Zn+5.1


21

preparation of dioctylzinc, see Chapter 6.


8. Confirmed by GCMS analysis. Nickel catalysed conjugate reduction of ",$-unsaturated ketones is aknown process, see: Caporusso, A.M.; Giacomelli, G.; Lardicci, L. J. Org. Chem. 1982 , 47, 4640. Seealso reference 3c.

9. Kindly pr ovided by Prof. J. Martens. For synthesis, see: Behnen, W.; Mehler, T.; Martens, J .Tetrahedron: Asymmetry 1993 , 4, 1413.

10. a) Ullenius, C.; Christenson, B. Pure Appl. Chem. 1988 , 60, 57. b) Krause, N.; Wagner, R.; Gerold, A.J. Am. Chem. Soc. 1994 , 116, 381.

11. For a disc ussion about ligand accelerated asymmetric catalysis, see: Berrisford, D.J.; Bolm, C. ;Sharpless, K.B. Angew. Chem. Int. Ed. Engl. 1995 , 34, 1059.

12. Van Koten, G.; No ltes, J.G. in Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F.G.A.;Abel, E.W., Eds.; Pergamon: Oxford, 1982 , Vol. 2, Chapter 14.

13. Alexakis, A.; Frutos, J.; Mangeney, P. Tetrahedron: Asymmetry 1993 , 4, 2427.

14. Alexakis, A.; Frutos, J.C.; Mangeney, P. Tetrahedron: Asymmetry 1993 , 4, 2431.

15. Hiemstra, H.; Wynberg, H. Tetrahedron Lett. 1977 , 2183.

16. Hathaway, B.J. In Comprehensive Coordination Chemistry; Wilkinson, G.; Gillard, R.D.; McLeverty,J.A., Eds.; Pergamon: Oxford, 1987 , Vol. 5, Chapter 53.

17. a) House, H.O. Fischer, Jr., W.F. J. Org. Chem. 1968 , 33, 949. b) Corey, E.J.; Beames, D.J. J. Am. Chem.Soc. 1972 , 94, 7210. c) M. Suzuki, T. Suzuki, T. Kawagishi, Y. Morita, R. Noyori, Isr. J. Chem. 1984 ,24, 118. d) A. Alexakis, S. Mutti, J. F. Normant, J. Am. Chem. Soc. 1991 , 113, 6332.

18. Recently enantioselective 1,4-additions of Grignard reagents to cyclic enones, catalysed by a CuI /chiral bidentate phoshine complex, has been reported: M. Kanai, K. Tomioka, Tetrahedron Lett. 1995 ,36, 4275.

19. Organophosphorus ligands have played a dominant role in the development of synthetic methodologyfor the preparation of chiral products by asymmetric hydrogenation using transition metal catalysts;for an extensive review, see: Morrison, J.D. Asymmetric Synthesis, Chiral Catalysis; Academic Press:New York, 1985 , Vol. 5.

20. R. Hulst, N. K. de Vries, B. L. Feringa, Tetrahedron: Asymmetry 1994 , 5, 699.

21. For a review see: Beaucage, S.L.; Iyer, R.P. Tetrahedron 1992 , 48, 2223.

22. a) Gagnaire, D.; Robert, J.B.; Verrier, J. Bull. Soc. Chim. Fr. 1968 , 6, 2392. b) Mosbo, J.A.; Verkade,J.G. J. Am. Chem. Soc. 1973 , 95, 4659. c) Schiff, D.E.; Richardson, Jr., J.W.; Jacobson, R.A.; Cowley,A.H., Lasch, J.; Verkade, J.G. Inorg. Chem. 1984 , 23, 3373 and references therein.

Chapter 5

22

23. a) Pastor, S.D.; Hyun, J.L.; Odorisio, P.A.; Rodebaugh, R.K. J. Am. Chem. Soc. 1988 , 110, 6547 andreferences therein. b) van Rooy, A. Rhodium Catalysed Hydroformylation with Bulky Phosphites asModifying Ligands, Ph.D. Thesis, University of Amsterdam, 1995 , Chapter 7.

24. Posner, G.; Frye, L.L. Isr. J. Chem. 1984 , 24, 88.

25. a) Leyendecker, F.; Laucher, D. Tetrahedron Lett. 1983 , 24, 3517. b) Leyendecker, F.; Laucher, D .Nouv. J. Chim. 1985 , 9, 13.

26. See Section 2.3.

27. a) Hof, R.P.; Poelert, M.A.; Peper, N.C.M.W.; Kellogg, R.M. Tetrahedron: Asymmetry 1994 , 5, 31. b)Fitzpatrick, K.; Hulst, R.; Kellogg, R.M. Tetrahedron: Asymmetry 1995 , 6, 1861.

28. Hoffmann, E.W.; Kuchen, W.; Poll, W.; Wunderlich, H. Angew. Chem., Int. Ed. Engl. 1979 , 18, 415.

29. Fabbri, D.; Delogu, G.; De Lucchi, O. Tetrahedron: Asymmetry 1993 , 4, 1591.

30. a) Hof, R.P. Enantioselective Synthesis and (Bio)catalysis, Ph.D. Thesis, University of Groningen, 1995 .b) de Vries, A.H.M.; Hof, R.P.; Staal, D.; Kellogg, R.M.; Feringa, B.L. Submitted to Tetrahedron:Asymmetry.

31. Kindly provided by Charon Zondervan. Synthesis according to: Frost, C.G.; Williams, J.M.J .Tetrahedron Lett. 1993 , 34, 2015.

32. a) Posner, G.H. Org. React. 1972 , 19, 1. b) Lipshutz, B.H. Synthesis, 1987 , 325.

33. Rossiter, B.E.; Swingle, N.M. Chem. Rev. 1992 , 92, 771. For catalytic enantioselective conjugat eadditions of organocuprates, see Section 2.3.

34. Khar asch, M.S.; Reinmuth, O. Grignard Reactions of Nonmetallic Substances, Prentice-Hall, Ne wYork, 1954 .

35. For a discussion on uncatalysed 1,4-addition of Grignard reagents, see Jansen, J.F.G.A. Stereoselective1,4-Additions, Ph.D. Thesis, University of Groningen, 1991 .

36. Cahiez, G.; Alami, M. Tetrahedron Lett. 1989 , 30, 3541.

37. Arai, M.; Lipshutz, B.H.; Nakamura, E. Tetrahedron 1992 , 48, 5709.

38. Rück, K.; Kunz, H. Synthesis, 1993 , 1018.

39. For transition metal catalysed vinyl and alkynyl transfer, see: a) Hooz, J.; Layton, R.B. J. Am. Chem.Soc. 1971 , 93, 7320. b) Bernardy, K.F.; Floyd, M.B.; Poletto, J.F.; Weiss, M.J. J. Org. Chem. 1979 , 44,1438. c) Schwartz, J.; Carr, D.B.; Hansen, R.T.; Dayrit, F.M. J. Org. Chem. 1980 , 45, 3053. d) Lipshutz,B.H.; Dimock, S.H. J. Org. Chem. 1991 , 56, 5761. e) Wipf, P.; Smitrovich, J.H.; Moon, C.-W. J. Org.Chem. 1992 , 57, 3178.

40. For trimethylaluminium additions, see: a) Ashby, E.C.; Noding, S.A. J. Org. Chem. 1979 , 44, 4792. b)Fuijwara, J.; Fukutani, Y.; Hasegawa, M.; Maruoka, K.; Yamamoto, H. J. Am. Chem. Soc. 1984 , 106,5004. c) Westermann, J.; Nickisch, K. Angew. Chem., Int. Ed. Engl. 1993 , 32, 1368. d) Kabbara, J. ,Flemming, S.; Nickisch, K.; Neh, H.; Westermann, J. Tetrahedron 1995 , 51, 743 and references therein.

41. The high potential to develop a enantioselective version of this selective copper(I) catalysed alky ltransfer from TM A to enones has very recently also been noticed by another research group: Takemoto,Y.; Kuraoka, S.; Hamaue, N.; Iwata, C. Tetrahedron: Asymmetry 1996 , 7, 993.They have reported a CuOTf catalysed 1,4-addition of TMA to 3,4,4-trimethylcyclohexa-2,5-dienone(5.19) with enantioselectivities up to 68% in the presence of 20 mol% of oxazoline 5.21 and 120 mol%of tert-butyldimethylsilyl triflate (TBDMSOTf).

O

5.19

+ TMA5.21 (20 mol%)CuOTf (5 mol%)

TBDMSOTfTHF, 0°C

O

88% yield68% e.e.

5.20

N

O

O

O

5.21


23

OO

P N

6.1

toluene-10°C

6.1 (20 mol%)CuI (10 mol%)

+ Et2Zn

Et2Zn+toluene-10°C

6.1 (20 mol%)CuI (10 mol%)

Et

O

*

PhPh

O

Ph Ph

O

*

O

47% e.e.

35% e.e.

6.4 6.5

6.36.2

105

Chapter 6

Novel Chiral Monodentate Phosphorus Amidites; Synthesis and Application as Ligands in the

Copper Catalysed Conjugate Addition ofDialkylzinc Reagents to Cyclic and Acyclic Enones

6.1 Introduction

As described in th e previous chapter a new catalytic system capable of enantioselectiveconjugate addition of diethylzinc to both a cyclic and acyclic substrate has bee ndeveloped. The novel trivalent phosphorus amidite 6.1 showed a remarkable ligand-acceleration in the CuI catalysed conjugate addition of diethylzinc to cyclohexenone(6.2 ) an d chalcone ( 6.4) resulting in an enantioselective alkyl transfer to bot hsubstrates (Scheme 6.1). Although the enantioselectivities are moderate, thes epreliminary experiments (see Section 5.3) provide several approaches for furthe rinvestigation, which will be described in the present chapter.

Scheme 6.1

Chapter 6

106

6.2 Determination of the molecular structure of the complex of CuI and 6.1

In order to examine the structure of the catalytically active complex (or the precursor)in th e solid state crystallisation experiments were undertaken. When CuI and tw oequi valents of ligand 6.1 were crystallised from toluene transparent block-shape dcrystals of the c opper-ligand complex were obtained, but determination of the structureby X-ray failed due to rapid loss of included toluene molecules. With benzene a ssolvent the process of deterioration of the crystals was slower and suitable colourlessneedles were obtained. The molecular structure is shown in Figure 6.1 and selecte dbond distances and angles are given in Table 6.1.The geometry around the copper nucleus of the complex is a slightly distorte dtetrahedron. An unexpected number of three chiral ligands is bound to the coppe rcreating a C -symmetry with the fourth position occupied by the iodide. The Cu-I bond3

length and the Cu-P bond lengths resemble distances reported in the molecula rstructure of [PEt CuI] and a 1,3,2-oxazaphospholidine !CuCl complex, respectively.3 4

1 2

The P-N distance of 1.626(17) Å is in good agreement with the bond length found inthe molecular structure of the free ligand (1.636(3) Å) and reported values i n3

analogueous uncoord inated trivalent phosphorus amidites. The amine group in the free4

liga nd and in the coordinated ligand is planar. The planar stereochemistry of th e3

nitrog en, found in the solid state for many amino phosphorus compounds, wherei nphosph orus is tri- or tetracoordinate, has been attributed to the presence o felec tronegative substituents on phosphorus, such as oxygens, which may serve t oenhance B bonding from N to P. 4

Examinati on of the molecular structure showed that crucial positions for ligan dmodificati ons are the amine group and the 3- and 3'-positions of the binaphthyl part ofthe ligand. However, before we turned our attention to the synthesis of these modifiedligand s, the catalytic activity of the crystalline complex was examined. With 10 mol%of the complex, the addition of diethylzinc to cyclohexenone proceeded selectively toth e 1,4-product and 6.3 was isolated in 76% yield with an e.e. of 32% (reactio nconditions as in Scheme 6.1), which are comparable results with those obtained withthe in situ prepared chiral catalyst.

Novel Chiral Phosphorus Amidites; Synthesis and Application in the Copper Catalysed Conjugate Addition

107

Table 6.1 Selected interatomic distances (Å) and bond angles (deg.) of (6.1) CCuI.3

Cu(1) - I(1) 2.640(3) P(1) - O(1) 1.639(18)Cu(1) - P(1) 2.258(7) P(1) - O(2) 1.665(17)Cu(1) - P(2) 2.275(7) P(1) - N(1) 1.626(17)`Cu(1) - P(3) 2.234(8) N(1) - C(21) 1.45(3)

N(1) - C(22) 1.43(4)

I(1) - Cu(1) - P(1) 106.4(2) O(1) - P(1) - N(1) 108.2(9)I(1) - Cu(1) - P(2) 104.6(2) O(2) - P(1) - N(1) 95.5(9)I(1) - Cu(1) - P(3) 103.5(2) O(1) - P(1) - O(2) 102.2(9)P(1) - Cu(1) - P(2) 109.2(3) P(1) - O(1) - C(1) 118.2(15)P(1) - Cu(1) - P(3) 112.2(3) P(1) - O(2) - C(20) 116.4(14)P(2) - Cu(1) - P(3) 119.7(3) P(1) - N(1) - C(21) 124.9(17)Cu(1) - P(1) - O(1) 107.0(6) P(1) - N(1) - C(22) 118.1(16)Cu(1) - P(1) - O(2) 121.5(6) C(21) - N(1) - C(22) 113.0(19)Cu(1) - P(1) - N(1) 120.6(7)

PCl3Et3N-70°C

OHOH

HNR2Et3N-40°C

= OO

P NR

R

OO

P Cl

OO

P NR

R[

6.6 6.7

6.8 - 6.17


109

metal, causing strong metal-ligand bonds, which results in more ligands coordinatedto the metal. However, when the withdrawing capacity is too large (very high P-value)coordination of more than one ligand would induce such a positive charge on the metalresulting in instable metal complexes. 8

Phosp horus amidites [P(NR' )(OR) ] represent a class of trivalent phosphoru s2 2

compounds hardly recognized as ligands for catalytic transformations. Only recentlyachiral phosphorus amidites have been applied as modifying ligands in the rhodiumcat alysed hydroformylation of 1-octene and styrene. The P-value of the applie d8

phosphorus amidites (R = aryl, approximately 21) is lower than the P-value o f7,8

arylphosphites (. 31) and higher than the P-value of arylphosphines ( . 13), indicating8

a moderate electron withdrawing ability.Several novel phosphorus amidites, all derived from enantiomerically pure 2,2' -bina phthol, were synthesised (this Section) and examined as ligands in the coppe r9

catalysed conjugate addition of diethylzinc to cyclohexenone and chalcone (Section6.4).

Modifications at the amine moietyLigand 6.1 has been s ynthesised by reaction of ( S)-2,2'-binaphthol ( 6.6) and HMPT (seeEq. 5.2). However, for the synthesis of modified phosphorus amidites another route10

has to be developed due to the lack of availability of analogueous phosphoru stri amides. Starting from phosphoryl chloride 6.7 , prepared according to a literatur eprocedure, phosphorus amidites were synthesised by nucleophilic substitution with11

a wide v ariety of secondary amines (Scheme 6.2). The results of the successfu lsyntheses are compiled in Table 6.2.

Chapter 6

110

Scheme 6.2

With diethylamine, dibenzylamine, and piperidine the corresponding phosphoru samidites were synthesised in satisfactory isolated yields (entries 1, 3, and 4 ,respectively). However, when the sterically demanding di- i-propylamine and cis-2,6-dimethylpiperidine were employed, the yield of phosphorus amidite is decreasing toca. 30% (en tries 2 and 5, respectively). For compound 6.9 the yield was slightl yimproved to 38% by using lithium di- i-propylamine (LDA) instead of the amine ,whereas refluxin g of the reaction mixture for 4 h did not influence the yield of product.One attempt to prepare compound 6.9 via a reversed synthesis using Cl PN( i-Pr) and2 2

6.6 was not successful. This, however, seems to be a feasible route according t oreferences 4 and 8.Phosp horus amidites 6.8-6.12 were remarkably stable to air and moisture and wer epur ified by washing with a saturated aqueous solution of sodium carbonate and acolumn chromatog raphy (SiO , hexane/CH Cl ). Storage under an inert atmosphere was2 2 2

not necessary. This stability is possibly a result of the combination of two phenoli csubstituents and one amine substituent on phosphorus. For example, in contrast to theseligands, reaction of the phenyl member of TADDOL ( ",","',"'-tetraaryl-2,2-dimethyl-1,3 -dioxolane-4,5-dimethanol) with PCl and piperidine furnished the phosphoru s12

3

amidite I (Figure 6.2, phosphorus resonance at 131 ppm). However, isolation of thispr oduct from the reaction mixture did not succeed, using the techniques mentione dabove.A general trend tow ard decreased shielding at phosphorus (downfield shift) is observedin the P NMR spectra of 6.1 (148.7 ppm), 6.8 , and 6.9 (Table 6.2) with an increase in31

the steric bulk of the amine groups bound to phosphorus. This downfield shift of thephosphorus resonance has been attributed to a decrease of the s character in th ephosphorus lone pair, probably induced by an increased phosphorus-nitrogen bond7

lenght. In an attempt to expand this sequence, hexamethyldisilazane - the obviouslyexpected di-t-butylamine is not commercially available - was used in the procedur eoutlined in Scheme 6.2. Although P NMR of the reaction mixture revealed a n31

expected downfield shifted value (152.7 ppm), we were not able to isolate phosphorusamidite II (Figure 6.2) from this mixture, probably due to instability of this particularcompound. The trend to decreased shielding at phophorus with increased steric bulkof the amine group is also perceptible in compounds 6.11 and 6.12 . Unfortunately, theextension to 2,2,6,6- tetramethylpiperidine to afford phosphorus amidite III (Figure 6.2)was not successful. Only other phosphorus compounds, possibly phosphites, wer edetected by P NMR.31

A remark able high yield of phosphorus amidite 6.13 (81%, entry 6) was achieved with


111

the chiral (3 R, 4R)-3,4-diphenylpyrrolidine (solid material), possibly a result of the13

purity of the a mine employed. Except for entry 10, all other secondary amines (liquids)were purchased, d ried over potassium hydroxide, and used without further purification.Proba bly higher yields of phosphorus amidites will be achieved with distilled amines.With N-methylpiperazine, morpholine, and thiomorpholine we were able to introducea he tero atom in the amine group attached to phosphorus (entries 7-9). Again th eobtained pho sphorus amidites 6.14-6.16 were purified by column chromatography andwere stable to ai r and moisture. Compared to piperidine substituted phosphorus amidite6.11 no substantial shift of the phosphorus resonance was observed for 6.14-6.16 ,indicating a minor influence of the hetero atom in the amine group on the phosphoruslone pair.

diethylamine

dibenzylamine

piperidine

morpholine

thiomorpholine

di(2-methoxyphenyl)amine

-methylpiperazineN

1

2

3

5

4

6

7

8

9

10

[OO

P N

[OO

P N

[OO

P NPh

Ph

[OO

P NPh

Ph

[OO

P N O

[OO

P N S

[OO

P N N

[ NPOO

O

O

[OO

P N

[OO

P N

59 149.8

54 144.7

32 146.5

65 145.5

81 148.9

30 145.0

61 144.6

43 145.0

45 139.9

didi- -propylamine

(3 )-3,4-diphenylpyrrolidine, 4R R

-2,6-dimethylpiperidinecis

151.728 (38)

6.17

6.16

6.15

6.14

6.13

6.11

6.12

6.10

6.8

6.9

Chapter 6

112

Table 6.2 Synthesis of phosphorus amidites 6.8-6.17 from phosphoryl chloride 6.7and accessory P NMR chemical shifts.31 a

entry secondary amine employed product yield (%)* (ppm)b c

VIV

OO

P N N POO

Ph Ph

OO

P N

O

IIIIII

O

Ph

O

Ph Ph

Ph

O

OP N

OO

P NOO

P NSiMe3

SiMe3


113

a. Reacti on conditions, see Scheme 6.2 and text. b. Isolated yields. c. In CDCl . d. With lithium di- i-3

propylamine instead of the amine.

Figure 6.2 Structures of phoshorus amidites I-V.

Wit h di(2-methoxyphenyl)amine a phosphorus amidite was prepared with oxyge n14

atoms present at a position capable of forming a chelated six-membered ring with thephosph orus and the coordinated copper. Since the product 6.17 was not stable t oco lumn chromatography (SiO hexane/toluene), isolation of 6.17 was performed b y2

precipitation fr om the reaction mixture with hexane. Compared to all the other amiditesthe phosphorus resonance of this aryl substituted amidite is shifted upfield indicatinga increased shielding of the phosphorus.Fu rthermore, with sym-dimethylethylenediamine the bidentate ligand 6.18 wasprepare d quite easily and isolated in 45% yield. The phosphorus resonance o fbisamidite 6.18 (148.9 ppm) is nearly the same as observed for the monodentat eanalogueou s amidite 6.1 (148.7 ppm), indicating the same shielding of the phosphorusin amidites 6.1 and 6.18 .

OO

P N NOO

P

6.18

Chapter 6

114

In an attempt to synthesise a bidentate phosphorus-oxygen ligand, 2-pyrrolidinone wasused in t he procedure outlined in Scheme 6.2. However, the initially forme dphosphorus amidite IV (Figure 6.2) - a phosphorus resonance at 140.7 ppm wa sobserved - is extremely sensitive to oxygen and / or moisture and only oxidize dproduc ts could be isolated using the isolation procedures mentioned above. Mos tprobably the carbonyl group reduces the stability of the amidite (oxidation o rhydro lysis of the P-N bond). In further investigations the use of, for example ,di(methoxyethyl)amine seems to be a feasible option to develop a stable phosphorusamidite with other coordinating groups present as well.Unfortunately, the reaction of diphenylhydrazine with two equivalents of phosphorylchloride 6.7 was also not successful. Again the presumably formed amidite V (Figure6.2) cou ld not be isolated from the reaction mixture using the isolation technique smentioned above.

Modifications at the 3,3'-position of the binaphthyl part of the ligandExaminatio n of the molecular structure in Section 6.2 showed that the 3- and 3' -positions o f the binaphthyl part of the ligand are crucial for ligand modification. Since3,3' -disubstituted binaphthols have shown highly promising catalytic activity fo rasymmetric induction in a wide range of reactions a lot of effort have been devoted15

to the synthesis of such compounds. Cram and co-workers prepared series o fsymmetri cal 3,3'-disubstituted binaphthols mainly via Mannich intermediates and by16

Grignard cross coupling of the key 3,3'-dibromo-2,2'dimethoxy-binaphthyl. A more17

direc t and versatile method for 3,3'-functionalisation of 2,2'-binaphthol, using 2,2' -oxygen-based directed metalation groups (for example the methoxymethyl grou p(MOM)), has been reported by Snieckus and co-workers. 18

Here we report our efforts on the synthesis of 3,3'-disubstituted binaphthols and theirconversion to the corresponding phosphorus amidites (Scheme 6.3). According to aliterat ure procedure (S)-2,2'-binaphthol ( 6.6) was methylated to compound 6.19 ,17

successively dilithiated, and quenched with bromine to give compound 6.21 in 48%yield (lit. 65%) after column chromatography (SiO , hexane/CH Cl /EtOAc 50:2:1).17

2 2 2

Th e generated dilithium compound of 6.19 was also quenched with methyl iodid efurnis hing compound 6.20 in 63% isolated yield. For comparison, with the MO M


115

protected analogueous compound this ortho lithiation-methyl iodide quenchin grea ction was performed in 82% yield. (In our hands this MOM-directed lithiation ,18

methyl io dide quenching, and deprotection with MeOH (saturated with HCl) furnishedcompound 6.23 in 70% overall yield.) For the preparation of 3,3'-diphenyl binaphthol,the dibromo compound 6.21 was treated with phenyl boronic acid under Suzuki crosscouplin g conditions to furnish 6.22 in 95% isolated yield, which is much higher than19

via the Grignard cross coupling procedure. 17

The 2,2'dimethoxy-3,3'-disubstituted derivatives 6.20-6.22 were deprotecte dquan titatively (BBr in CH Cl ) into the corresponding binaphthols 6.23-6.25 , and3 2 2

17

suc cessively converted to the corresponding phosphorus amidites 6.26-6.28 withHMPA in toluene (yield > 75%). Comparison of the phosphorus resonance of amidite 6.26 (145.9 ppm) with th eunsubstituted analogous amidite 6.1 (148.7 ppm) revealed an increased shielding atphosphorus (upfield shift) contrary to the observed trend of shifting the phosphorusresonance to low fiel d with increasing the size of substituents on phosphorus. Probably7

the electronic effe cts of the substituents on the 3,3'-position have more influence on thephosphorus resonance. The trend observed in the resonance values found for 6.27(148.8 ppm ), 6.28 (147.4 ppm), and 6.26 can be explained by the presence of electron-withdrawing groups (Br and Ph) and a electron-donating (Me) group, respectively.

OROR

OROR

OROR

Br

Br

OROR

Ph

Ph

6.6 R = H6.19 R = Me

a

b, c b, d

e

6.20 R = Me6.23 R = H

f6.21 R = Me6.24 R = H

f6.22 R = Me6.25 R = H

f

OO

Br

Br

P NOO

Ph

Ph

P NR'

R'NP

OO

g g g

6.26 R' = Me 6.27 6.28

6.29 R' = i-Pr6.30 R' = -(CH 2CH2)2O

Chapter 6

116

Scheme 6.3 Synthesis of 3,3'-substituted binaphthols and their conversion to thecorresponding phosphorus amidites. (a) methyl iodide (excess), K CO ,2 3

acetone (98%) (b) n-BuLi (2.2 equivalent), TMEDA, diethyl ether (c)methyl iodide, diethyl ether (d) bromine, diethyl ether (e) phenylboronicacid, Pd(PPh ) (0.06 equivalent), dimethoxyethane, NaHCO , water (f)3 4 3

BBr , CH Cl (g) HMPT, toluene.3 2 2

According to the reaction sequence of Scheme 6.2, the 3,3'-dimethyl binaphthol 6.23was converted with PCl , and successively LDA and morpholine to the corresponding3


117

phosphorus amidites 6.29 and 6.30 in 37% and 51% yield, respectively. The shifts ofthe phospho rus resonances of 6.29 (149.2 ppm) and 6.30 (142.1 ppm) are in goo dagreement with those determined for the unsubstituted amidites 6.9 and 6.15 .

6.4 Examination of the modified phosphorus amidites as chiral ligands inthe copper catalysed conjugate addition of diethylzinc tocyclohexenone and chalcone

With the ligands described in the previous Section the effect of the modified ligandson the e nantioselectivity in the copper catalysed ethyl transfer from diethylzinc t ocyclohexenone ( 6.2) and chalcone ( 6.4) were examined (the two model reactions, seeScheme 6.1). When ligands 6.8 , 6.10 , 6.11 , 6.15 , 6.18 , and 6.26 were tested in the CuIcatalysed ad dition of diethylzinc to 6.2 the reaction is slower (2 days at -10 EC) and theregioselectivity to the 1,4-product 6.3 is less selective (ca. 80%), compared to the CuI/ 6.1 catalys ed conjugate addition. Yields and e.e. values found for 6.3 are much lower,as well . The important ligand-acceleration, observed for 6.1 (Section 5.3), seems to beabsent for the modif ied phosphorus amidites. This lack of efficient catalysis is probablydue to solubility problems of the in situ generated modified phosphorus amidite-coppercomplexes. Contrary to the catalysis with CuI and 6.1 , in all these cases the solutionsof the copper complexes, prepared prior to catalysis, were not clear.In order to enhance the solubility of the complexes, the CuOTf (OTf = triflate =trifluoromethanesulfonate) catalysed conjugate addition reaction was investigated. Itis known that the triflate counterion is fully dissociated in solution. For the CuOTf20

catalysed addition of diethylzinc to 6.2 a higher regioselectivity was observed [after30 h at r oom temperature complete conversion was achieved with ca. 75% 1,4 -selectivity (other detected compounds were reduced products)], compared to the CuIcatalysed reaction (S ection 5.3). With a catalytic amount of phosphorus amidite 6.1 (2.2equival ents compared to copper) again a ligand-acceleration was observed in th eCuOTf cata lysed conjugate addition of diethylzinc to 6.2 , as well as to 6.4 (Eq. 6.1 and6.2). High isolated yields of 1,4-products 6.3 and 6.5 (82% and 91%, respectively)21

were obtained with e.e.'s slightly higher (e.e. 39% for 6.3) and significantly higher (e.e.65% for 6.5) than those determined for the CuI / 6.1 catalysed reactions (see Scheme6.1). Fu rthermore, the addition reaction shown in Eq. 6.1 appeared to proceed with thesame regio- and enantioselectivity in CH Cl as in toluene. So, CH Cl can be used as2 2 2 2

well to dissolve the complexes prepared from CuOTf and the modified phosphoru samidites. For further variations of the reaction conditions, see Section 6.5.With this knowledge the effect of the modified phosphorus amidites ( 6.8-6.18 and 6.26-

Et2Zn+toluene, -10°C

CuOTf (3-5 mol%)L* (6-11 mol%)

Et2Zn+toluene, -10°C

CuOTf (3-5 mol%)L* (6-11 mol%)

Et

O

*

PhPh

O

Ph Ph

O

*

O

6.4 6.5

6.2 6.3

(6.1)

(6.2)

Chapter 6

118

6.30) on the enantioselectivity of the ethyl transfer from diethylzinc to the two modelsubstrates could be e xamined (Eq. 6.1 and 6.2). The results are summarised in Table 6.3.

In general a solution of a catalytic amount of (CuOTf) Cbenzene (3-5 mol%) and chiral2

l igand (L*, 2.2 equivalents compared to copper) in toluene (3-5 ml) was stirred a tambient tempera ture for 1 h under argon furnishing a clear solution. If the solution wasnot clear, appropriate amounts of CH Cl were added. Substrate was added (1.0-2. 02 2

mmol), th e mixture was cooled to -20 EC and diethylzinc in toluene (1.1 M, 1. 5equival ent) was added. Stirring was continued at -10 EC for 16 h. An aliquot of th esolution (0.1 ml) was taken and quenched with aqueous 1 N HCl (1 ml) and th econve rsion was determined with GC analysis. In all cases regioselectivities to the 1,4-products were higher than 90% at > 95% conversion. The 1,4-products were isolatedand purified by column chromatography (yields > 75%), prior to e.e. determination toav oid problems and mistakes. Control experiments have elucidated that thi spurification procedure did not affect the e.e. value.Except for 6.14 , all phosphorus amidites induce asymmetry in the ethyl transfer fromdiethylzinc to enones. Contrary to all other cases, the ligand 6.14 could not be detectedby TLC after the reaction. Besides the 1,4-product only 2,2'-binaphthol was isolated,probably formed b y nucleophilic attack of the tertiar amine to the trivalent phosphorus.For the other ligands the following trends were observed:- Sterically demanding substituents on the amine group have improved the e.e.

values found for 6.3 and 6.5 up to 60% and 83%, respectively (entry 2, see alsoentry 4).

- Modifications on the amine group resulting in phosphorus amidites with a nincreased shielding at phosphorus (upfield shift, see Section 6.3) showed in allcases an en hanced enantioselectivity for the addition to 6.2 , compared to ligand6.1 (entries 3, 5, and 7-9). However, an alternating effect is observed for th e


119

addition to 6.4 . Ligands with substituents containing hetero atoms on the aminegroup, retain the ceiling e.e.'s found for 6.5 (entries 7 and 8), whereas ligandswith aromatic groups in the amine part furnished 6.5 with lower e.e.'s (entries 3and 5) or showed to be detrimental for enantioselective ethyl transfer to 6.4(entry 9).

- Employing ph osphorus amidites with methyl groups on the 3- and 3'-position ofthe ligand have only effect on the enantioselective alkyl transfer to 6.2 (e.e. isincreased) and 6.4 (e.e. is decreased) if the amine group is small (entry 11). Withdi-i-propylam ine or morpholine as amine group comparable enantioselectivitiesare observed as found with the unsubstituted analogueous amidites (compareentries 14 and 15 with entries 2 and 7).

- Phenyl or bromide groups on the 3- and 3'-position of the ligand seems to bedetriment al to the enantioselective ethyl transfer to 6.4 , whereas still good e.e.'sare found for the conjugate addition to 6.2 (entries 12 and 13).

Table 6.3 Enantioselective conjugate addition of diethylzinc to 6.2 and 6.4catalysed by CuOTf and phosphorus amidites 6.8-6.18 and 6.26-6.30. a

entry chiral ligand e.e. of 6.3 (%) ( S) e.e. of 6.5 (%) ( R)b c

1 6.8 27 - 2 6.9 60 83 3 6.10 53 53 4 6.12 43 79 5 6.13 47 50 6 6.14 0 -d

7 6.15 50 71 8 6.16 55 70d

9 6.17 48 13d

10 6.18 37 -d

11 6.26 56 52 12 6.27 51 23 13 6.28 35 < 20 e

14 6.29 59 81 15 6.30 51 76

a. Reaction conditions, see text and Eq. 6.1 and 6.2. b. Enantiomeric excess of 3-ethyl cyclohexanone ( 6.3) det-ermined by derivati sation with optically pure 1,2-diphenylethylenediamine. Comparison of the optical rotation22

Chapter 6

120

of 6.3 with known data gave the absolute configuration. c. Enantiomeric excess of 1,3-diphenylpentan-1-one23

(6.5) determined by HPLC analysis: Daicel, Chiralcel OD; 0.25% i-PrOH in hexane, flow rate 1.0 ml/min, UVdetector (2 54 nm). Comparison of retention times of ( R)- and ( S)-6.5 with known data gave the absolut econfiguration. d. Appropriate amount of CH Cl (1-3 ml) was added to the catalyst solution, see also text. e.24

2 2

Not base line separated, due to unknown reason.

Fu rthermore, employing the bidentate phosphorus amidite 6.18 (1.1 equivalen tcompared to copper) in the CuOTf catalysed ethyl transfer to 6.2 furnished the 1,4 -product 6.3 in compar able yield and e.e. (entry 10) as was found with the correspondingmo nodentate ligand 6.1 (2.2 equivalents compared to copper). This indicates th epre sence of two chiral monodentate ligands in the catalytically active complex o raggregate ( vide infra).

6.5 Variation of the reaction conditions, substrate, and dialkylzinc reagent

Variation of reaction conditionsBefore the most enantioselective ligand 6.9 was employed in the copper catalyse dconjugate addition to other substrates or with other dialkylzinc reagents, the reactioncondition s were examined in more detail. Until this moment still 2.2 equivalents of thechiral monodentate ligand compared to the amount of CuOTf was used to prepare thechiral cat alyst in a solution of toluene at -10 EC. This was based on the results obtainedwith ligand 6.1 and CuI (see Section 5.3). In order to optimise the conditions for theCuOTf catalysed conjugate addition we have studied some variables.First, the addition of diethylzinc to 6.2 , catalysed by CuOTf (3 mol%) and 6.1 (6.5mol%), was performed at an initially temperature of -60 EC. After 16 h the temperaturewas increased to -15 EC, the 1,4-product 6.3 was isolated (75% yield), and the e.e .determination revealed a value of 15%. (Reaction at room temperature revealed th esame values as found for reaction at -10 EC, yield 79%, e.e. 34%). Secondly, th eaddition r eaction employed with 5 equivalents of chiral ligand 6.1 compared to CuOTf(3 mol%) app eared to proceed only at room temperature. After 10 days the 1,4-product6.3 was isolated (45% yield), but the e.e. was very low (< 10%, see also Section 5.3).Probably , with 5 equivalents of chiral ligand the copper ion is not accessible fo rsubstrate and reagent.Furth ermore, when the CuOTf / 6.9 catalysed addition of diethylzinc to 6.4 wasperforme d with a catalyst solution, which was filtered prior to reaction, the e.e .observed for 6.5 dropped to 63% (instead of 83%). The undissolved miniscule particlesseems to be essentia l. Also a slow addition of the reagent (1 h) to the mixture of catalyst(CuO Tf / 6.9) and substrate ( 6.4) - often essential for high enantioselectivity in th eco pper catalysed conjugate addition of Grignard reagents, see Section 2.3 - had a


121

negative effect on the enantioselectivity (e.e. 71%), whereas an inversed addition ofreagent and substrate had no effect on yield or e.e. (yield 88%, e.e. 82%). Presumably,the actual chiral catalyst is only formed when all reagents are present. Finally , when the conjugate addition reactions of diethylzinc to 6.2 and 6.4 wer eperformed with a catalyst derived from Cu(OTf) (2 mol%) and 6.9 (2.2 equivalent s2

compared to copper) the reactions are even faster and slightly improved e.e.'s for 6.3and 6.5 were observed (63% and 87%, respectively, Table 6.4, entries 1 and 2) .Remarkably, t his Cu catalysed addition furnished product 6.5 with an even higher e.e.II

(90%) when performed at a lower temperature (-50 EC).In copp er catalysed Diels-Alder reactions better results were obtained using th ecounterio n SbF . When we performed the Cu catalysed conjugate addition o f- 20 II

6

diethylzinc to 6.2 and 6.4 with Cu(SbF ) and ligand 6.9 , only reasonable conversion6 2

to the 1,4-product (ca. 80%) was observed for the cyclic substrate. Although the e.e.observed for 6.3 was higher (71%, Table 6.4) this seems to be not the catalyst of choice.The rea ction is slower and less regioselective than the Cu(OTf) catalysed version and2

with chalcone no 1,4-product could be isolated, due to unknown reasons.The experiments described in this Section are only performed once, however, they allclearl y illustrate the complicated character of this catalytic addition reaction. I nco nclusion, the best chiral catalyst so far is prepared from Cu(OTf) and tw o2

equivalents of the chiral ligand (preferably 6.9) in toluene and / or CH Cl . Cooling to2 2

ca. -20EC, followed by fast addition of substrate and reagent (in arbitrarily order) willfurnish the 1,4-product with the highest e.e. values. Two remarks have to be made: (1)The actual chiral catalyst when Cu(OTf) is used, is probably a Cu species formed by2

I

in situ reduction of the Cu complex. (2) In general the copper catalysed conjugateII 25

addition reaction is completed within 3 h. For practical reasons most reactions were runfor 16 h, however, control experiments revealed that this had no influence on yield ore.e. of the 1,4-products.

Variation of substrateTh e copper catalysed conjugate addition of diethylzinc to various substrates wer eperformed with the optimised catalyst, derived from Cu(OTf) and 2.2 equivalents of2

phosphorus amidite 6.9 . The results are shown in Table 6.4. The e.e.'s of all 3 -substituted cyclic ketones were determined by derivatisation with optically pure 1,2-diphenylethylene diamine. Enantiomeric excesses of the acyclic ketones 6.38 , 6.40 ,22

and 6.42 were determined by HPLC analysis (Daicel, Chiralcel OD or OJ column).

6.33

6.3

*

6.366.356.346.2 n = 1

6.31 n = 06.32 n = 2

OOOO

*( )n

O

( )n

O

O

6.43 6.44 6.45

RPh

O

Ph R

O

*

O

EtO

O

OEt

Ph

O

EtO

O

OEt

Ph*

6.4 R = Ph6.37 R = Me6.39 R = 4-OCH3Ph6.41 R = 2-pyridyl

6.5 6.38 6.40 6.42

Chapter 6

122

Table 6.4 Addition of diethylzinc to various ",$-unsaturated compounds catalysedby Cu(OTf) and chiral phosphorus amidite 6.9.2

a

entry substrate 1,4-product yield (%) e.e. (%)b c

1 6.2 6.3 78 63 (71) d

2 6.4 6.5 84 87 (90) e

3 6.31 - - - 4 6.32 6.33 76 55 5 6.34 6.35 76 81 6 6.36 - - - 7 6.37 6.38 75 - 60 f

8 6.39 6.40 85 80 9 6.41 6.42 69 2910 6.43 6.44 88 3511 6.45 - - -

a. Reaction con ditions as in Eq. 6.1 and 6.2. b. Isolated yields of pure compounds. c. E.e determination, see textand / or experimental section. d. With Cu(SbF ) , see reference 20 for procedure. e. Temperature -50 EC. f.6 2

Not base line separated.

Compared to the reaction with 2-cyclohexen-1-one (entry 1) remarkable results were

(6.3)

6.46

O

O

( )3

1) HBEt22) Et2Zn

) 2O

O

Zn ( )3

(


123

observed with 2-cyc lopenten-1-one ( 6.31 ) and 2-cyclohepten-1-one ( 6.32 ) as substrate.With 6.32 compar able yield and e.e. for 1,4-product 6.33 were found (entry 4), whereaswith 6.31 hardly any 1,4-product could be detected by GC analysis, possibly due todim er- and oligomerisation of the enone. Isolation of the product from the reactio nmixture was not possible. Oligomerisation of 6.31 and 6.2 catalysed by early transitionmetals is a known process, with a much higher activity for 6.31 .26

Steric ally demanding methyl groups on the 4-position of the enone enhanced th eenant ioselectivity to 81% (entry 5), while with isophorone ( 6.36 ) no addition reactionoccurred, probably due to steric hindrance.The acyclic enones benzylideneacetone ( 6.37 ) and 4'-methoxychalcone ( 6.39 ) wer ealkylated in good yields and with enantioselectivities of 60% and 80%, respectively(entries 7 and 8). The pyridyl substituted chalcone 6.41 gave the 1,4-product 6.42 in27

only 29% e.e. (entry 9), presumably due to competitive binding of the copper catalystto the pyridine group of the enone ( vide infra).Conjugate addition to less reactive ",$-unsaturated esters is possible with compound6.43 fur nishing 6.44 in high yield (entry 10) with a moderate e.e. (35%). Sinc edeterm ination of the e.e. of product 6.44 failed with chiral HPLC and GC methods, thee.e. was determined by chiral HPLC analysis after reduction of both ester groups withLiAlH to the corresponding diol. With cyclic ester 6.45 no addition reaction occurred,4

nor with enh anced temperature or prolonged reaction times. Only starting material wasdetected by TLC and GC analysis. Diester 6.43 is probably activated by complexationof the chiral copper complex to both carbonyl groups. This could account for th emoderate asymmetric induction as well ( vide infra).

Variation of dialkylzinc reagentAs already mentioned in Section 3.5, diorganozinc reagents are highly preferred a salkyl donor for catalytic enantioselective alkylation since they can be prepared fromrel atively cheap alkenes by the elegant boron-zinc exchange process developed b yKnoch el and co-workers. A second important advantage is that they tolerate th e28

presence of a wid e range of functional groups. In Section 3.5 we have used dioctylzinc,prepa red from 1-octene, successfully in the nickel / chiral amino alcohol catalyse daddition to chalcone (50% yield, 72% e.e.). In this Section preliminary experiment swith a functionalised dialkylzinc reagent will be described.

3

O

O

O

( )*3

O

O

O

( )* *3

O

OPhPh

O ( )

6.47 6.48 6.49

Chapter 6

124

Ther efore, 4-penten-1-yl-acetate was hydroborated with diethylborane an dsuc cessively transmetalated with pure diethylzinc furnishing the functionalise ddialkylzinc reagent 6.46 (Eq. 6.3). After evaporation of the volatiles, the formation of6.46 was established by iodolysis of the dialkylzinc reagent followed by GC analysis(c a 75%). This reagent was used as such in the conjugate addition reaction to thre edifferent substrates (6.2 , 6.4 , and 6.34 ) catalysed by Cu(OTf) (1 mol%) and phosphorus2

amidite 6.9 (2.2 equivalents compared to copper) in toluene at -15 EC for 48 h. In allthree cases the corresponding 1,4-products 6.47 , 6.48 , and 6.49 were isolated in fairlygood yiel ds (ca. 50%) and with e.e.'s of 56%, 62%, and 65%, respectively (Figure 6.3),which a re slightly lower as those found for the analogueous addition with diethylzinc.Traces of ethyl transfer to the enones were detected, due to incomplete evaporation ofthe volatil es. So, if the reagent is purified before use, yields of the 1,4-products will beimproved probably.

Figure 6.3 1,4-Products obtained via conjugate addition of functionaliseddialkylzinc reagent 6.46 catalysed by Cu(OTf) and phosphorus amidite2

6.9.

6.6 Mechanistic aspects

Although no quantitative measurements and kinetic studies were performed an ddetailed examinations of the complexes, formed in solution, by NMR studies ar elacking, an intermedi ate which could account for the enantioselective alkyl transfer willbe proposed. The following observations support this proposal:(1) The copper catalyst derived from bidentate ligand 6.18 and CuOTf (ratio of 1.1 to1) showed th e same results as found for the corresponding monodentate ligand 6.1 andCuOTf (ratio of 2.2 to 1) in the addition reaction of diethylzinc to 6.2 . This indicatesthat during reaction two monodentate ligands are bound to the copper ion. (Fo rsimplicity the catalyst is presented as a monomeric unit, however, it is quite feasiblethat this unit is part of a dinuclear complex or a larger aggregate.) The presence of atleast two ligands during reaction was also established by the examination of th e

0 25 50 75 1000

20

40

60

80

e.e.

of 6

.5 (%

)

e.e. of 6.9 (%)


125

relationship betw een the e.e. of ligand 6.9 and the e.e. of the product 6.5 . Scalemic 6.9 29

with defined e .e. was employed in the Cu(OTf) catalysed addition of diethylzinc to 6.42

resulting in a negative nonlinear relationship (Figure 6.4). Nonlinear relationship sbetween the e.e.'s of chiral auxiliaries and e.e.'s of products have been interpreted interms of differences in the chemical behaviour of diastereomeric complexes. When30

a nonli near relationship is observed the role of at least two ligands in the catalyti cprocess is assumed. See also Section 3.6 for a related comprehensive discussion.

Figure 6.4 The effect of the e.e of 6.9 on the e.e. of 6.5 in the Cu(OTf) catalysed2

addition of diethylzinc to 6.4.

(2) The conjugate addition reaction of diethylzinc to the pyridyl substituted chalcone6.41 and diester 6.43 furnished the corresponding 1,4-products with moderat eenantiomeric excesses. In these cases the possible coordination of the chiral coppercomplex to the carbon-carbon double bond is probably competiting with coordinationto other groups - the pyridine entity and the 1,3-diester functionality, respectively -present in the substrates, resulting in a less enantioselective alkyl transfer.We therefore propose the intermediates and pathways responsible for th een antioselective addition as follows (Figure 6.5). First an ethyl fragment will b etransferred to the copper complex. (Transmetalation of organozinc compounds into thecorresponding organocopper reagents is generally accepted. ) The reacting substrate31

coordinates t o the chiral copper entity by B-complexation of the carbon-carbon doublebon d and is activated by the Lewis acidic zinc center. The small electronegativit y32

R - 6.5

Ph Ph

O

Ph

PhO

L

LCu

ZnX

ZnX

Cu

L

L O

6.2S - 6.3

O +EtZnX

L2Cu

O

VI

VII

Ph Ph

O

6.4

L2CuX (or L2CuX2) + Et2Zn

Chapter 6

126

value of z inc (1.6) relative to that of copper (1.9) suggest the generation of a zin cenolat e instead of a copper enolate. The remaining sites in the tetrahedra l33

coordination sphere of the copper ion are occupied by two phosporus amidite sgenerating a chiral environment with one favourable B-face selective alkyl transfer. Viachiral intermediates VI and VII (with the same chiral coordination sphere) productswith the S configur ation will be formed for s- trans enones ( i.e. 6.2 , see intermediate VI)and products with the R configuration for s- cis enones ( i.e. 6.4 , see intermediate VII ).

Figure 6.5 Intermediates and pathways in the copper catalysed alkyl transfer toenones which could account for the enantioselectivity and theconfiguration of the products ( L = chiral phosphorus amidite).

The use of very small quantities of catalysts allowed relatively clean formation of the


127

zinc enolate wh ich suggests the possibility to trap the enolates by electrophiles leadingto products with two or three (in the case of prochiral electrophiles) stereogeni ccenters. However, preliminary quenching reactions of the zinc enolate of 6.2 or 6.4 ,34

generated from the s ubstrate and diethylzinc in the presence of 1 mol% of Cu(OTf) and2

2 mol% o f 6.9 in toluene at -25 EC, with methyl iodide did not result in the tande mreactio n. Only 'normal' 1,4-products were isolated even at higher reaction temperatureor in the presen ce of one equivalent of lithium chloride (the lithium enolate is probablymore active ), possibly due to the insufficient reactivity of alkyl iodides. Very recently,zinc enolates, pr epared from cyclic enones by copper catalysed addition of diethylzinc,we re successfully trapped with aldehydes and allyl acetates. An enantioselectiv e21

version with ligand 6.9 seems to be feasible.

6.7 Summary and future prospects

In this Chapter th e successful synthesis of a wide variety of novel phosphorus amidites,all derived from enantiomerically pure 2,2'-binaphthol, has been described. Th emodification s of the amine group and at the 3- and 3'-position in the binaphthol part ofthe liga nd were based on the determination of the molecular structure of the Cu Icomplex of ligand 6.1 . CuOTf was successfully employed to increase the solubility ofthe in situ formed chiral complexes. Application of the novel phosphorus amidites inth e CuOTf catalysed conjugate addition of diethylzinc to both a cyclic and acyli csubstrate (cycloh exenone and chalcone) revealed an enhancement in enantioselectivityfor the ethyl transfer to both model substrates. The highest e.e. values were observedfor the di- i-propylamine derived amidite, whereas sterically demanding groups on the3- and 3'-position were detrimental for high e.e.'s. Further remarkable features of theCuOTf / phosphorus amidite catalysed conjugate addition reaction are: (1) the ver yeffective ligand- acceleration, (2) excellent regioselectivity to nearly pure 1,4-products,and (3) the efficiency of a monodentate ligand in asymmetric catalysis. Furthermore,with this catalytic system all kinds of ",$-unsaturated ketones can be successfull yal kylated at the 3-position in high yields and with moderate to high e.e.'s. The onl ylimitations seems to be the reactivity of the substrate and the presence of competitivecoordinating groups.Another advantage of this alkyl transfer via dialkylzinc reagents is the possibility tost art with relatively cheap alkenes and the tolerance of these reagents to functiona lgr oups. Preliminary experiments showed that dioctylzinc (see Chapter 3) and afunctionalised dialkylzinc reagent, prepared directly from the corresponding alkenes,can be successfu lly employed. The 1,4-products with e.e. values comparable with those

Chapter 6

128

observed for diethyl zinc were isolated in moderate yields. Optimisation of the yield andextension to other functional groups are most likely feasible.A model is proposed which could account for the enantioselective ethyl transfer to bothcyclic and acyclic substrates and the observed configuration of the 1,4-products. Thismodel is ba sed on the observed nonlinear effect, the results of the ethyl transfe remploying a bident ate phosphorus amidite, and the moderate enantioselectivities foundfor substrates with competitive coordinating groups. In principle, the generated zincenolate might be trapped with an electrophile creating a product with three stereogeniccenter s. The mechanism of the copper catalysed conjugate addition is still unexploredand the oxidat ion state of the active copper complex is not clear yet. NMR experimentsof the co mplexes in solution should eluicidate these unknown reaction characteristics.


GeneralAll reactions described in this Chapter were performed under argon using flame-driedsta ndard Schlenk equipment. Toluene was distilled from sodium and stored unde rnitrogen. Di chloromethane (CH Cl ) was distilled from P O and stored under nitrogen.2 2 2 5

P NMR spectra were recorded on a Varian Gemini 200 at 80.95 MHz. For all other31

general remarks, see Section 3.8.

MaterialsTriethylamine (Et N) was freshly distilled and stored over potassium hydroxide. All3

secon dary amines were purchased from Aldrich or Acros and dried with potassiu mhyd roxide unless stated otherwise. The following compounds were commerciall yavailable and used without purification: PCl (Merck), tetramethylethylenediamin e3

(TMEDA, Acros), bromine (Aldrich), phenyl boronic acid (Acros), BBr (Merck), 2-3

cyclopenten-1- one ( 6.31 ; Aldrich), 2-cyclohepten-1-one ( 6.32 ; Aldrich), 4,4-dimethyl-2-cyc lohexen-1-one (6.34 ; Aldrich), isophorone ( 6.36 ; Acros), benzylideneaceton e(6.3 7; Acros), diethylbenzalmalonate ( 6.43 ; Aldrich), 5,6-dihydro-2 H-pyran-2-on e(6.45 ; Aldrich), 4-penten-1-yl-acetate (Aldrich), BH CMe S (Aldrich), and BEt3 2 3

(Aldrich). (CuOTf) Cbenzene (90%) and Cu(OTf) were both purchased from Aldrich.2 2

(CuOTf) Cbenzene was stored under nitrogen and Cu(OTf) was dried under vacuum2 2

before use.Pure diethylzinc was purchased from Aldrich, cannulated to Schlenck equipment andstor ed under argon atmosphere. Caution: diethylzinc is highly pyrophoric and verysensitive to moisture.


129

(3 R, 4R)-3,4-Diphenylpyrrolidine, di(2-methoxyphenyl)amine, 1-(2-pyridyl)-3 -13 14

phenyl-2-propen-1-one ( 6.41 ), and Pd(PPh ) were kindly provided by T. Vries, R.27 353 4

La Crois, S. Otto, and A.M. Schoevaars, respectively.For all other materials, see Section 3.8.

X-ray crystallography of (6.1) CCCuI3

Colourless transparent block-shaped crystals were obtained when CuI and tw oequivalents of ligand 6.1 were dissolved in toluene but X-ray structure determinationfailed by very quick evaporation of the included toluene molecules leading t odeter ioration of the crystals. With benzene as solvent the process of deterioration wasslower and a parallelepiped colourless crystal having approximate dimensions of 0.12x 0.1 2 x 0.44 mm was used for data collection by an Enraf-Nonius CAD-4 Fdiffractometer interfaced to a MicroVAX-2000 computer (performed by A. Meetsma,Universi ty of Groningen). Crystal data and experimental details of the structur edetermination are compiled in Table 6.5.

Table 6.5 Crystal data, data collection, structure solution, and refinement for(6.1) CCuI3

Crystal dataChemical formula C H N O P CCuIC2(C H )66 54 3 6 3 6 6

Formula weight (g.mol ) 1424.77-1

Crystal system monoclinicSpace group P2 2 21 1 1

a (Å) 15.525(2)b (Å) 19.957(2)c (Å) 24.535(3)V (Å ) 9999.9(5)3

Z 4D (g.cm ) 1.245calc

-3

F(000) 2920µ(Mo K ¡) (cm ) 8.07-1

Approx. crystal dimension (mm) 0.12 x 0.12 x 0.44

Data collectionRadiation Mo K" (0.71073 Å)Monochromator graphiteTemperature (K) 1301 range ( E) 1.02-26.5Total data 8605

Chapter 6

130

Unique data 8569Observed data 5371 ( I $ 2.5 F(I))

RefinementNumber of reflections 5354Number of refined parameters 405Final agreement factors:R = G( # # F # - # F # #)/G# F # 0.120F 0 C 0

wR = [G(w( # F # - # F # ) )/Gw# F # ] 0.1310 C 02 2 1/2

The cell dimensions were obtained by least-squares refinement of the settings anglesof 22 r eflections with 13.25 E < 2 < 18.85 E, from various parts of reciprocal space. Theunit ce ll was identified as monoclinic, space group P2 2 2 . The experimental stability1 1 1

was checked by t hree standard reflections. These reference reflections, measured everythree hours, indicated a large linear increase of 25% over 188.7 h of X-ray exposuretime. A 360 E R-scan for a reflection close to axial (241) showed a variation in intensityof less than 16% about the mean value. Intensity data were corrected for Lorentz andpolarisation effects, scale variation, for absorption with DIFABS and reduced to F .0 36

The s tructure was solved by Patterson methods and extension of the model wa saccomplishe d by direct methods applied to difference structure factors using th eprogram DIRDIF. The positional and anisotropic thermal displacement parameters for37

the non-hydrogen atoms were refined with block-diagonal least-squares procedure s(CRYLSQ) minimizing the function Q = G [ w( # F # - # F # ) ], where the weight w ish 0 C

2

defined as 1/F (F) and F and F are the observed and calculated structure facto r20 C

amplitudes, r espectively. Subsequent Fourier summations showed density which couldbe correlated to t he two disordered benzene solvent molecules, but showed also densitywhich could not be correlated to disordered solvent molecules. No of discrete modelcould be fitted in this density. The potential solvent area volume in the unit cell is 1410Å (in two regions), which means 19% of the unit cell. High thermal displacemen t3

motion was sited for some carbon positions and some converged to non-positiv edefin ite thermal displacement parameters when allowed to vary anisotropically ,suggesting some degree of disorder, which is in line with the observation of loosingsolvent molecules. The two benzene molecules were included in the refinement as rigidgroups. The residual electron densities were of no chemical significance.The hydrogen atoms were included in the final refinement riding on their carrier atomswith their positions calculated by using sp or sp hybridisation at the C-atom a s2 3

appropri ate with a fixed C-H distance of 0.98 Å. Final refinement on F by full-matrix0

lea st-squares techniques with anisotropic thermal displacement parameters for th enon-carbon, non-hydrogen atoms, isotropic thermal displacement parameters for the


131

carbon atoms and o ne common thermal displacement parameter for the hydrogen atomsconverged at R = 0.120 ( wR = 0.131). Refinement converged at the high R value dueF

to severe disorder. Weights were introduced in the final refinement cycles. A fe wreflections with ( w( # # F # - # F # # ) > 30) were excluded from the final refinement0 C

cycle. Owing to the presence of highly disordered molecules within the holes of thecry stal lattice, several significant peaks, not exceeding 2.2 e/Å , in the last residua l3

Fourier maps were found, but their geometry could not be rationalised. Neutral atomscattering factors were used and anomalous dispersion factors were included in F . AllC

calculati ons were carried out on the HP9000/735 computer at the University o fGroningen with the program packages X tal, PLATON (calculation of geometric data)38 39

and a locally modified version of the program PLUTO (preparation of illustrations).40

Elemental analysisThe parallelepiped colourless crystals, obtained by slow evaporation of benzene andused for crysta l structure determination, were also submitted to elemental analysis. Theincorporated benzene molecules were partly evaporated. Anal. calcd for C H N O P CCuI: C, 62.5; H, 4.3; N, 3.3; Cu, 5.0. Found: C, 63.6; H, 4.5;66 54 3 6 3

N, 3.1; Cu, 4.9.

General procedure for the synthesis of ligands 6.8-6.17 To a cooled solution (-60 EC) of PCl (270 µl, 3.0 mmol), Et N (860 µl, 6.0 mmol) and3 3

toluene ( 5 ml) was added a warm solution (60 EC) of ( S)-2,2'-binaphthol ( 6.6) (860 mg,3.0 mmol) and toluene (25 ml) in 5 min. After stirring for 2 h the reaction mixture waswarmed to room temperature and filtered under argon atmosphere. The filtrate wa streated with Et N (410 µL, 2.9 mmol) and 2.9 mmol of the corresponding secondary3

amine at -40 EC. After 16 h at ambient temperature, the reaction mixture was filtered,concentrated and purified by chromatography (SiO , hexane:CH Cl 2:1) to give the2 2 2

pure amidite as a colourless amorph compound. Stripping with CH Cl furnished the2 2

phosporus amidites as foamy solids with still some solvent molecules incorporated .Crystallisation from diethyl ether / CH Cl mixtures should give crystalline material.2 2

Yields are given in Table 6.2. The physical data of the compounds are as follows:

O,O'-(1,1'-Dinaphthyl-2,2'-diyl)-N,N-diethylphosphorus amidite (6.8)["] + 501 E (c 0.50, CH Cl ). H NMR * 1.08 (t, J = 7.2 Hz, 6H), 2.79-3.21 (m, 4H),2 0 1

D 2 2

7.18- 7.58 (m, 8H), 7.89-8.05 (m, 4H). C NMR * 14.01 (q), 44.22 (t), 121.97 (d) ,13

122.14 (d), 122 .91 (s), 123.98 (s), 124.11 (s), 124.41 (d), 124.65 (d), 126.07 (d), 126.93(d), 128.20 (d), 129.85 (d), 130.11 (d), 130.65 (s), 131.34 (s), 132.50 (s), 149.17 (s),149.69 (s). P NMR * 149.8. HRMS calcd for C H NO P: 387.139, found 387.139.31

24 22 2

Chapter 6

132

O,O'-(1,1'-Dinaphthyl-2,2'-diyl)-N,N-di-i-propylphosphorus amidite (6.9)["] + 591 E (c 0.68, CHCl ). H NMR * 1.19 (d, J = 6.84 Hz, 6H), 1.24 (d, J = 6.84 Hz,2 0 1

D 3

6H), 3.37 (heptet, J = 6.84 Hz, 1H), 3.42 (heptet, J = 6.84 Hz, 1H), 7.55-7.22 (m, 8H),7.99- 7.89 (m, 4H). C NMR * 24.27 (q), 24.43 (q), 44.48 (d), 44.74 (d), 121.69 (s) ,13

122.26 (d), 122.32 (d), 123.91 (s), 124.10 (d), 124.45 (d), 125.63 (d), 125.72 (d), 126.96(d), 128.04 (d), 128.10 (d), 129.18 (d), 130.00 (d), 130.32 (s), 131.16 (s), 132.54 (s),132.67 (s), 150.03 (s), 150.22 (s), 150.35 (s). P NMR * 151.7. HRMS calcd fo r31

C H NO P: 415.170, found 415.170.26 26 2

O,O'-(1,1'-Dinaphthyl-2,2'-diyl)-N,N-dibenzylphosphorus amidite (6.10)["] + 163E (c 0.54, CH Cl ). H NMR * 3.47 (AB system, J = 15.0 Hz, 1H), 3.53 (AB2 0 1

D 2 2

system, J = 15.0 Hz, 1H), 4.25 (AB system, J = 15.0 Hz, 1H), 4.29 (AB system, J = 15.0Hz, 1H), 7.16-7.51 (m, 17H), 7.68-7.88 (m, 3H), 7.98 (d, J = 8.1 Hz, 1H), 8.07 (d, J =8.65 H z, 1H). C NMR * 47.90 (t), 48.33 (t), 121.48 (d), 122.16 (d), 122.90 (s), 123.6313

(s), 123.85 (s), 124.62 (d), 124.86 (d), 126.10 (d), 126.91 (d), 127.06 (d), 127.35 (d),128.21 (d), 128.4 0 (d), 128.84 (d), 130.15 (d), 130.31 (d), 130.64 (s), 131.24 (s), 132.48(s), 137.83 (s), 149.22 (s), 149.80 (s). P NMR * 144.7. HRMS calcd for C H NO P:31

34 26 2

511.170, found 511.170.

O,O'-(1,1'-Dinaphthyl-2,2'-diyl)-N-(1-piperidinyl)phosphorus amidite (6.11)H NMR * 1.36-1.61 (m, 6H), 2.88-3.11 (m, 4H), 7.21- 7.56 (m, 8H), 7.83-8.00 (m, 4H).1

C NMR * 24.88 (t), 26.89 (t), 26.97 (t), 45.10 (t), 45.50 (t), 122.08 (d), 122.18 (d) ,13

122.74 (s), 123 .93 (s), 124.08 (s), 124.45 (d), 124.68 (d), 125.97 (d), 126.95 (d), 128.25(d), 129.75 (d), 130.17 (d), 130.65 (s), 131.33 (s), 132.58 (s), 149.27 (s), 149.90 (s). P31

NMR * 145.5. HRMS calcd for C H NO P: 400.147, found 400.147.25 23 2

O,O'-(1,1'-Dinaphthyl-2,2'-diyl)-N-[1-(2,6-dimethyl)piperidinyl]phosphorus amidite(6.12)["] + 472 E (c 0.76, CH Cl ). H NMR * 1.17 (d, J = 7.1 Hz, 3H), 1.30 (d, J = 7.1 Hz,2 0 1

D 2 2

3H), 1 .46-1.91 (m, 6H), 3.42-3.52 (m, 1H), 3.86-3.98 (m, 1H), 7.21-7.58 (m, 8H), 7.90-8.01 (m, 4H). C NMR * 14.43 (t), 23.61 (q), 23.96 (q), 30.99 (t), 31.42 (t), 45.62 (d),13

45.81 (d), 46.27 (d), 47.01 (d), 121.89 (s), 122.36 (d), 124.30 (d), 124.62 (d), 125.80 (d),125.92 (d), 127.0 9 (d), 128.21 (d), 129.32 (d), 130.15 (d), 130.54 (s), 131.31 (s), 132.65(s), 132.72 (s), 149.64 (s), 150.02 (s). P NMR * 146.5. HRMS calcd for C H NO P:31

27 26 2

427.170, found 427.170.


133

O , O ' - ( 1 , 1 ' - D i n a p h t h y l - 2 , 2 ' - d i y l ) - N - [ 1 - ( ( 3 R , 4 R ) - 3 , 4 -diphenyl)pyrrolidinyl]phosphorus amidite (6.13)["] + 164 E (c 1.35, CH Cl ). H NMR * 3.39-3.62 (m, 6H), 7.16-7.34 (m, 12H), 7.43-2 0 1

D 2 2

7.51 (m, 4H), 7. 64 (t, J = 8.8 Hz, 2H), 7.96-8.08 (m, 4H). C NMR * 52.27 (d), 52.31 (s), 13

53.19 (t), 53.40 ( t), 121.52 (d), 121.73 (d), 122.83 (s), 123.77 (s), 123.83 (s), 124.48 (d),124.61 (d), 125.14 (d ), 125.92 (d), 125.99 (d), 126.73 (d), 126.83 (d), 128.06 (d), 128.12(d), 128.17 (d), 128.89 (d), 129.79 (d), 130.11 (d), 130.61 (s), 131.18 (s), 132.46 (s),132.65 (s) , 139.99 (s), 149.55 (s), 149.86 (s) 149.92 (s). P NMR * 148.9. HRMS calcd31

for C H NO P: 537.186, found 537.186.36 28 2

O,O'-(1,1'-Dinaphthyl-2,2'-diyl)-N-(1-(N-methyl)piperazidinyl)phosphorus amidite(6.14)H NMR * 2.15-2.26 (m, 7H), 2.91-3.11 (m, 4H), 7.11-7.58 (m, 8H), 7.85-8.03 (m, 4H).1

C NMR * 44.31 (t), 44.67 (t), 47.28 (q), 56.54 (t), 56.71 (t), 121.93 (d), 123.06 (s),13

123.97 (s), 124.14 ( s), 124.86 (d), 125.01 (s), 126120 (d), 126.82 (d), 128.33 (d), 131.13(d), 130.53 ( d), 130.81 (s), 131.44 (s), 132.34 (s), 132.92 (s), 149.26 (s), 150.01 (s). P31

NMR * 145.0. HRMS calcd for C H N O P: 414.150, found 414.150.25 23 2 2

O,O'-(1,1'-Dinaphthyl-2,2'-diyl)-N-(1-morpholinyl)phosphorus amidite (6.15)["] + 364E (c 0.19, CH Cl ). H NMR * 2.93-3.21 (m, 4H), 3.53-3.61 (m, 4H), 7.24-7.6020 1

D 2 2

(m, 8H), 7.92-8.04 (m, 4H). C NMR * 44.37 (t), 44.73 (t), 67.78 (t), 67.87 (t), 121.9013

(d), 122.87 (s), 123.83 (s), 123.95 (s), 124.75 (d), 124.90 (s), 126.20 (d), 126.92 (d),128.37 (d), 130.0 9 (d), 130.41 (d), 130.78 (s), 131.42 (s), 132.34 (s), 132.96 (s), 149.32(s), 149.80 (s) . P NMR * 144.6. HRMS calcd for C H NO P: 401.118, found 401.118.31

24 20 3

O,O'-(1,1'-Dinaphthyl-2,2'-diyl)-N-(1-thiomorpholinyl)phosphorus amidite (6.16)["] + 383E (c 0.39, CH Cl ). H NMR * 2.41-2.53 (m, 4H), 3.20-3.38 (m, 4H), 7.22-7.5820 1

D 2 2

(m, 8H), 7.91-8.03 (m, 4H). C NMR * 42.53 (t), 42.71 (t), 44.41 (t), 44.77 (t), 122.0213

(d), 122.64 (s), 123.86 (s), 123.91 (s), 124.65 (d), 124.89 (s), 126.12 (d), 126.92 (d),128.31 (d), 130.1 0 (d), 130.43 (d), 130.77 (s), 131.40 (s), 132.41 (s), 132.83 (s), 149.43(s), 149.88 (s), 150.04 (s). P NMR * 145.0. HRMS calcd for C H NO PS: 417.095,31

24 20 2

found 417.095.

O,O'-(1,1'-Dinaphthyl-2,2'-diyl)-N,N-di(2-methoxyphenyl)phosphorus amidite (6.17)H NMR * 3.80 (s, 6H), 6.51 (dt, J = 7.7 Hz, J = 1.3 Hz, 2H), 6.80 (dd, J = 8.1 Hz, J =1

1.3 Hz, 2 H), 6.96 (dt, J = 7.7 Hz, J = 1.3 Hz, 2H), 7.12-7.43 (m, 9H), 7.49-7.57 (m, 2H),7.75 (d, J = 8.1 Hz, 1H), 7.93 (t, J = 9.4 Hz, 2H). C NMR * 55.82 (q), 112.24 (d) ,13

120.14 (d), 121.6 2 (s), 122.13 (d), 124.05 (s), 124.62 (d), 125.38 (d), 125.86 (d), 126.30

Chapter 6

134

(d), 126.74 (d), 127.06 (d), 127.73 (d), 128.22 (d), 128.52 (d), 130.05 (d), 130.22 (s),130.53 (d), 130.62 (d), 131.31 (s), 132.75 (s), 132.97 (s), 155.28 (s). P NMR * 139.9.31

HRMS calcd for C H NO P: 543.160, found 543.160.34 26 4

N , N ' - B i s [ O , O ' - ( 1 , 1 ' - d i n a p h t h y l - 2 , 2 ' - d i y l ) - N - m e t h y l p h o s p h o r usamidite]ethylenediamine (6.18)The gene ral procedure as described above was applied with 1.50 mmol of sym-dimethylethylenediamine. Yield 45%. H NMR * 2.37 (s, 3H), 2.40 (s, 3H), 3.01-3.191

(m, 2H), 3.29-3.42 (m, 2H), 7.24-7.62 (m, 16H), 7.90-8.04 (m, 8H). C NMR * 32.21 (q),13

47.09 (t), 47. 85 (t), 122.06 (d), 122.26 (d), 122.47 (s), 123.93 (s), 124.05 (s) 124.59 (d),124.81 (d), 126.09 (d ), 126.91 (d), 127.04 (d), 128.27 (d), 128.34 (d), 130.11 (d), 130.30(d), 13 0.72 (s), 131.38 (s), 132.56 (s), 132.83 (s), 149.45 (s), 149.98 (s), 150.10 (s). P31

NMR * 148.9. HRMS calcd for C H N O P : 716.199, found 716.199.44 34 2 4 2

Improved synthesis of compound 6.9The general procedure as described above was applied to prepare the phosphory lchloride 6.7 . This mixture was filtered in a preformed cooled (-40 EC) solution of LDA(2.8 mmol di- i-propylamine and 2.8 mmol n-BuLi) in THF (10 ml). After 16 h a tambient temperature, the reaction mixture was filtered over Celite, concentrated andpurifie d by chromatography (SiO , hexane:CH Cl 2:1) to give the pure amidite .2 2 2

Racemic- 6.9 was prepared according to this procedure.

3,3'-Dimethyl-2,2'-dimethoxy-1,1'-dinaphthyl (6.20)A sol ution of ( S)-6.19 (0.72 g, 2.29 mmol), prepared according to a publishe dprocedure, and TMEDA (1.81 ml, 12.0 mmol) in 40 ml of diethyl ether was cooled17

with ice/water. A 2.0 M solution of n-BuLi in hexanes (4.9 ml, 9.8 mmol) was addeddropwise over a period of 25 min. The mixture was stirred at 0 EC for 30 min and wasthen slow ly warmed to reflux. After refluxing for 20 h the resulting orange suspensionwas cooled to 0 EC and MeI (1.56 ml, 25 mmol) was added dropwise over a period of 45min resulting in a white suspension. After stirring for 16 h at ambient temperature thewhite mixture was poured into 200 ml of aqueous 1 N HCl. The mixture was extractedwith CHCl (250 ml) and the organic layer was washed with a saturated solution o f3

NaHCO (200 ml), brine (200 ml), dried (Na SO ) filtered, and evaporated to give the3 2 4

crude product as a yellow solid (0.94 g). The crude product was filtered over a shortcol umn (SiO , hexane:EtOAc:CH Cl 10:1:1) and recrystallised from hexane:EtOA c2 2 2

(10:1, 35 ml) to give 6.20 (0.46 g, 63%) as colourless crystals. H NMR * 2.57 (s, 6H),1

3.33 (s, 6H), 7.11-7.41 (m, 6H), 7.79-7.86 (m, 4H). C NMR * 17.23 (q), 60.07 (q) ,13

124.54 (d), 124.6 0 (s), 125.29 (d), 125.72 (d), 127.12 (d), 129.56 (d), 130.71 (s), 131.56


135

(s), 133.11 (s), 155.74 (s).

3,3'-Diphenyl-2,2'-dimethoxy-1,1'-dinaphthyl (6.22)A mixtur e of ( S)-6.21 (1.04 g, 2.20 mmol), prepared according to a publishe dprocedure, and Pd(PPh ) (0.153 g, 0.132 mmol) in 10 ml of dimethoxyethane was17 35

3 4

stirred at roo m temperature for 30 min. After addition of PhB(OH) (0.59 g, 4.84 mmol)2

and NaHCO (1.1 g in 13 ml of H O) the suspension was refluxed for 16 h resulting in3 2

a colour less solution standing above a brownish gum. After cooling to roo mtemperature the mixture was extracted with CH Cl (2 x 75 ml) and the organic layer2 2

was was hed with brine (25 ml), dried (Na SO ) filtered, and evaporated to give th e2 4

crude prod uct as a foam (1.22 g). The crude product was purified by colum nchromatograp hy (SiO , hexane:toluene 1:1) to give 6.22 (0.977 g, 95%) as a colourless2

powder. H NMR * 3.21 (s, 6H), 7.21-7.53 (m, 12H), 7.80 (d, J = 6.8 Hz, 4 H), 7.92-8.011

(m, 4H).

Deprotection of compounds 6.20, 6.21, and 6.22All 3,3'-disubstituted-2,2'-dimethoxy-1,1'-dinaphthyls were deprotected with BBr 3

according to a lit erature procedure furnishing the corresponding diols ( 6.23 , 6.24 , and17

6.25) in high yields (> 90%). Spectroscopic data (NMR and [ "] ) of compounds 6.232 0D

and 6.25 were in good agreement with data in the literature. 16-18

3,3'-Dibromo-2,2'-dihydroxy-1,1'-dinaphthyl (6.24)["] - 89E (c 0.58, CH OH). H NMR * 5.58 (s, 2H), 7.08 (d, J = 7.6 Hz, 2H), 7.23-7.4620 1

D 3

(m, 4H), 7.81 (d, J = 7.7 Hz, 2H), 8.23 (s, 2H). C NMR (CD OD / CDCl ) * 116.65 (s),133 3

119.26 (s), 128 .24 (d), 128.44 (d), 130.94 (d), 133.40 (s), 136.47 (d), 136.84 (s), 152.78(s).

General procedure for the synthesis of compounds 6.26, 6.27, and 6.28To a m ixture of a 3,3'-disubstituted-2,2'-dihydroxy-1,1'dinaphthyl (1 mmol) in 5 ml oftoluene hexamethylphosphorus triamide (1.5 mmol) was added. The mixture becameclear and aft er 1 h a white solid precipitated. The mixture was stirred for 3 h at ambienttemperature, 3 ml of diethyl ether was added, and a white solid precipitated. The solidwas collected by filtration, washed with diethyl ether (3 ml), and dried in vacuo to giveth e pure phosphorus amidites 6.26 -6.28 . Yields > 75%. The physical data of th ecompounds are as follows:

O,O'-(1,1'-Dinaphthyl-2,2'-diyl-3,3'-dimethyl)-N,N-dimethylphosphorus amidite(6.26)

Chapter 6

136

["] + 554E (c 0.14, CH Cl ). H NMR * 2.51 (s, 6H), 2.60 (d, J = 9.4 Hz, 6H), 7.12-7.4320 1D 2 2

(m, 6H), 7.73-7.89 (m, 4H). P NMR * 145.9. HRMS calcd for C H NO P: 387.139,3124 22 2

found 387.139.

O,O'-(1,1'-Dinaphthyl-2,2'-diyl-3,3'-dibromide)-N,N-dimethylphosphorus amidite(6.27)["] + 554 E (c 0.54, CH Cl ). H NMR * 2.59 (d, J = 9.4 Hz, 6H), 7.23-7.31 (m, 4H),2 0 1

D 2 2

7.41-7.49 (m, 2H), 7.84 (d, J = 7.7 Hz, 2H), 8.27 (d, J = 9.0 Hz). P NMR * 148.8 .31

HRMS calcd for C H Br NO P: 514.929, found 514.929.22 16 2 2

O,O'-(1,1'-Dinaphthyl-2,2'-diyl-3,3'-diphenyl)-N,N-dimethylphosphorus amidite(6.28)["] + 446 E (c 0.50, CH Cl ). H NMR * 1.96 (d, J = 9.4 Hz, 6H), 7.17-7.52 (m, 12H),2 0 1

D 2 2

7.70-7.82 (m, 4H), 7 .95-8.05 (m, 4H). C NMR * 34.01 (q), 34.40 (q), 124.32 (s), 124.7313

(d), 124.97 (d), 125.87 (d), 126.63 (d), 126.84 (d), 127.18 (d), 127.93 (d), 127.98 (d),128.11 (d), 128.21 (d ), 128.30 (d), 129.70 (d), 129.83 (d), 129.91 (d), 130.58 (d), 130.75(s), 130.95 (s), 132.21 (s), 132.39 (s), 134.12 (s), 134.88 (s), 138.03 (s), 147.28 (s). P31

NMR * 147.4. HRMS calcd for C H NO P: 511.170, found 511.170.34 26 2

Synthesis of compounds 6.29 and 6.30Compounds 6.29 and 6.30 were prepared according to the general procedure given forcompounds 6.8-6.17 using 3,3'-dimethyl-2,2'-dihydroxy-1,1'-dinaphthyl ( 6.23 ) and thecor responding amines (LDA (see improved synthesis for 6.9) and morpholine ,respectively). The physical data of the compounds are as follows:

O,O'-(1,1'-Dinaphthyl-2,2'-diyl-3,3'-dimethyl)-N,N-di-i-propylphosphorus amidite(6.29)Yield 37%. [ "] + 522 E (c 1.00, CH Cl ). H NMR * 1.19 (d, J = 4.8 Hz, 12H), 2.58 (s,2 0 1

D 2 2

3H), 2.60 (s, 3H), 3.32-3.40 (m, 2H), 7.12-7.38 (m, 6H), 7.75-7.82 (m, 4H). C NMR * 13

17.25 (q), 18.19 (q), 24.84 (q), 24.91 (q), 45.00 (d), 45.18 (d), 121.93 (s), 123.98 (s),124.18 (d), 124.44 (d ), 124.76 (d), 124.86 (d), 126.85 (d), 127.09 (d), 127.35 (d), 127.45(d), 129.03 (d), 129.50 (d), 130.28 (s), 130.35 (s), 131.12 (s), 131.70 (s), 149.41 (s),149.84 (s). P NMR * 149.2. HRMS calcd for C H NO P: 443.201, found 443.201. 31

28 30 2

O,O'-(1,1'-Dinaphthyl-2,2'-diyl-3,3'-dimethyl)-N-(1-morpholinyl)phosphorus amidite(6.30)Yield 51 %. ["] + 467 E (c 0.30, CH Cl ). H NMR * 2.50 (d, J = 9.4 Hz, 6H), 2.82-3.182 0 1

D 2 2

(m, 4H), 3.41-3.64 (m, 4H), 7.14-7.43 (m, 6H), 7.75-7.88 (m, 4H). P NMR * 142.1 .31


137

HRMS calcd for C H NO P: 429.149, found 429.149.26 24 3

Conjugate addition of diethylzinc to cyclohexenone (6.2) and chalcone (6.4) usingcatalytic amounts of (CuOTf) CCbenzene and chiral phosphorus amidites 6.8-6.18 and2

6.26-6.30 This p rocedure is typical for all conjugate addition reactions described in Section 6.4.A solutio n of (CuOTf) Cbenzene (3-5 mol%) and of chiral amidite (6-10 mol%) in 5 ml2

of toluene was stirr ed at ambient temperature for 1 h under argon. In general this resultsin a clea r solution. If not, appropriate amounts of CH Cl were added untill a clea r2 2

solution was o btained (see Table 6.3). Substrate was added (1.0-2.0 mmol), the mixturewas coo led to -20 EC and diethylzinc in toluene (1.1 M, 1.5 equivalent) was added .Stirrin g was continued at -10 EC for 16 h. An aliquot of the solution (0.1 ml) was takenand quenched w ith 1 ml of aqueous 1 N HCl. After extraction with 1 ml of diethyl etherthe con version was determined by GC analysis. Retention times (oven temperatur e100EC, flow 10 1 ml/min He): cyclo-2-hexen-1-one ( 6.2), 2.87 min; 3-ethylcyclohexan-1-one (6.3), 5.88 min; (oven temperature 225 EC, flow 101 ml/min He): chalcone ( 6.4),5.66 min; 1,3-diphenylpentan-1-one ( 6.5), 4.93 min. In all cases complete conversionwas achieved w ith regioselectivities to the 1,4-product > 90%. The mixture was pouredinto 25 ml of aqueous 1 N HCl and extracted with diethyl ether (3 x 20 ml). Th ecombined organic layers were washed with brine (25 ml), dried (MgSO ), filtered and4

evaporated to give the crude 1,4-products. (Caution: compound 6.3 is volatile and longevaporation times should be avoided.) After purification by column chromatography(SiO , hexan e:diethyl ether 5:1) the e.e.'s were determined. For 6.3 : Derivatisation with2

op tically pure 1,2-diphenylethylene diamine in CDCl (10 min, with some 4Å mo l3

sieves) followed by C NMR analysis. For 6.5 : HPLC analysis (see Section 3.8). H13 22 1

NMR and C NMR data of 6.3 and 6.5 were in good agreement with the data reported13

in Chapters 3, 4, and 5. E.e. values are given in Table 6.3.

Variation of reaction conditionsAl l these experiments, described in the first part of Section 6.5, were performe daccording to the general procedure with the modifications given in the text. Yields andenantioselectivities are also described in the text.

Conjugate addition of diethylzinc to various "",$$-unsaturated ketones employing achiral catalyst derived from Cu(OTf) and amidite 6.92

This procedure is typical for all substrates given in Table 6.4. A solution of Cu(OTf) 2

(2-3 mol%) and of chiral amidite 6.9 (4-6 mol%) in 5 ml of toluene was stirred a tambient temperature for 1 h under argon. Substrate was added (1.0-2.0 mmol), th e

Chapter 6

138

mixture was cooled to -20 EC and diethylzinc in toluene (1.1 M, 1.5 equivalent) wasadded. Stirring was continued at -10 EC for 16 h. The conversion was determined byTLC or G C analysis. When complete conversion was achieved the mixture was pouredinto 25 ml of aqueous 1 N HCl and extracted with diethyl ether (3 x 20 ml). If th econversion was not complete, longer reaction times and / or higher temperatures wereemployed , however, this did not result in any conversion (see text). The combine dorganic layers were washed with brine (25 ml), dried (MgSO ), filtered, and evaporated4

to give the crude 1,4-products. After purification by column chromatography (SiO ,2hexane:diethyl ether 5:1 unless stated otherwise) the e.e.'s were determined. Yields ande.e's are given in Table 6.4. Spectroscopic data (NMR) of products 6.38 and 6.40 werein good agreement with data in the literature. E.e.'s of 6.38 and 6.40 were determined24

by HPLC analysis; Daicel: Chiralcel OD, 1.0% iPrOH in hexane, flow rate 1.0 ml/min,UV detector (235.5 nm).

3-Ethylcyloheptan-1-one (6.33)H NMR * 0.89 (t, J = 7.0 Hz, 3H), 1.14-1.42 (m, 4H), 1.44-1.72 (m, 2H), 1.78-1.96 (m,1

4H). C NMR * 11.06 (q), 24.10 (t), 28.24 (t), 29.74 (t), 36.16 (t), 37.39 (d), 43.60 (t),13

49.31 (t), 214 .64 (s). E.e. determination according to the procedure of Alexakis and co-workers. Longer reaction time (16 h) for derivatisation is required (compared to 3-22

alkylcyclohexanone).

4,4-Dimethyl-3-ethylcyclohexan-1-one (6.35)H NMR * 0.84-1.07 (m, 1H), 0.85 (t, J = 6.8 Hz, 3H), 0.97 (s, 3H), 1.01 (s, 3H), 1.32-1

1.43 (m, 1H), 1. 55-1.73 (m, 3H), 1.94-2.07 (m, 1H), 2.21-2.49 (m, 3H). C NMR * 11.9313

(q), 19.22 (q), 23.03 (t), 28.46 (q), 32.67 (s), 38.08 (t), 40.25 (t), 42.03 (t), 48.60 (d),212.37 (s). E.e. determination, see 6.33 .

1-(2-Pyridyl)-3-phenylpentan-1-one (6.42)Purified by column chromatography (SiO , hexane:diethyl ether 2:1). H NMR * 0.812

1

(t, J = 7.4 Hz, 3H), 1.60-1.85 (m, 2H), 3.19-3.34 (m, 1H), 3.45-3.69 (m, 2H), 7.10-7.23(m, 1H), 7.24-7.47 (m, 4H), 7.40-7.47 (ddd, 1H), 7.74-7.82 (dt, 1H), 7.92-7.98 (dt, 1H),8.65-8.68 (ddd, 1H). C NMR * 11.79 (q), 29.32 (t), 42.42 (d), 43.98 (t), 121.66 (d),13

125.92 (d), 126.84 (d), 127.63 (d), 128.09 (d), 136.68 (d), 144.77 (s), 148.69 (d), 153.45(s), 200.70 (s). E.e. determined by HPLC analysis; Daicel Chiralcel OJ, 10.0% iPrOHin hexane, flow rate 1.0 ml/min, UV detector (240 nm), retention times 18.2 and 25.4min.

Diethyl 1-phenylpropylmalonate (6.44)


139

Purified by column chromatography (SiO , CH Cl :hexane 1:1). H NMR * 0.74 (t, J =2 2 21

7.3 Hz, 3H), 0.95 (t, J = 7.1 Hz, 3H), 1.31 (t, J = 7.1 Hz, 3H), 1.54-1.85 (m, 2H), 3.28 (dt,J = 10.9 Hz, J = 3.8 Hz, 1H), 3.66 (d, J = 10.9 Hz), 3.89 (q, J = 7.1 Hz), 4.25 (q, J = 7.1Hz), 7.13-7.33 (m, 5H). C NMR * 11.70 (q), 13.68 (q), 14.10 (q), 27.04 (t), 47.32 (d),13

58.68 (d), 61.0 5 (t), 61.47 (t), 125.26 (d), 126.80 (d), 128.21 (d), 128.39 (d), 129.00 (d),140.88 (s), 168.64 ( s). E.e. determination of 6.44 by chiral GC, chiral HPLC (Daicel: ODand OJ column), and by Eu-shift reagent failed, unfortunately. Therefore the diester6.44 was con verted to the corresponding diol by reduction with LiAlH in diethyl ether4

Yield 85%.1-hydroxy-2-hydroxymethyl-3-phenylpentane H NMR * 0.70 (t, J = 7.3 Hz, 3H), 1.48-1

1.69 (m, 1H), 1.76-1.97 (m, 2H), 2.50-2.62 (m, 1H), 2.9-3.5 (broad signal, 2H), 3.36-3.45 (m, 1H), 3.56-3.63 (m, 1H), 3.76-3.86 (m, 1H), 3.96 (dd, J = 10.9 Hz, J = 3.4 Hz,1H), 7.10-7.34 (m, 5H). C NMR * 12.16 (q), 25.90 (t), 45.51 (d), 46.61 (d), 64.32 (t),13

64.66 (t), 126.26 (d), 128.24 (d), 128.34 (d), 143.04 (s). E.e. determined by HPL Can alysis; Daicel Chiralcel OJ, 10.0% iPrOH in hexane, flow rate 0.5 ml/min, U Vdetector (208 nm), retention times 19.1 and 21.7 min.

Typical procedure for the conjugate addition of a functionalised alkene to enonescatalysed by Cu(OTf) and phosphorus amidite 6.9 (hydroboration-transmetalation-2

1,4-addition process)The fun ctionalised dialkylzinc reagents were prepared according to a procedur edevelope d by Knochel and co-workers. Here the procedure to form dialkylzin c28b

reagent 6.46 is described as well as the copper catalysed conjugate addition to threeenones leading to compounds 6.47-6.49 . 4-Penten-1-yl ace tate (2.81 ml, 20 mmol) was cooled with liquid nitrogen and degassedwith vacuum, than cooled with ice / H O and HBEt [21 mmol, 1.2 M, freshly prepared2 2

from BH CMe S (2M in diethyl ether) and BEt (1M in THF)] was added via a syringe3 2 3

(1 min). After 3 h stirring at room temperature an aliquot of the solution (0.1 ml) wastaken, o xidized with H O /NaOH, and analysed with GC (80% of the alcohol, retention2 2

times (oven temperature 110 EC, after 2 min 10 EC/min 6 160EC, flow 101 ml/min He):alkene 1.68 min; alcohol 4.78 min). The volatiles were removed under vacuum (0. 1mmHg, 0EC, 1 h) and the resulting organoborane was treated with pure Et Zn (2.2 ml,2

18 mmol) at 0 EC affording a dark coloured mixture. After stirring for 1 h at 0 EC and 1h at ambient temperature the excess of Et Zn and formed BEt were pumped off (0.12 3

mmHg, 0EC, 3 h). An aliquot of the residue (1 drop) was taken and added to a solutionof I in THF (dry). After 1 min the excess I was destroyed with Na S O in H O and the2 2 2 2 3 2

organic phase was analysed with GC (75% of 5-iodide-pentan-1-yl acetate, retentiontimes (oven temperature 110 EC, after 2 min 10 EC/min 6 200EC, flow 101 ml/min He):

Chapter 6

140

alkene 1.68 min; alcohol 4.78 min, iodide 5.62 min). The residue was diluted wit htoluene (5 ml) and added by syringe equipped with a filter (0.45 µm) to three cooled (-25EC) solutions of the chiral catalyst (Cu(OTf) and 6.9) and substrates 6.2 , 6.4 , and2

6.34 , respectively (see procedure above). For all three Schlenk flasks stirring wa scontinue d at -15 EC for 48 h. The conversions were monitored by TLC analysis and themixtures were poured into 25 ml of aqueous 1 N HCl and extracted with diethyl ether(3 x 20 m l). The combined organic layers were washed with brine (25 ml), drie d(MgSO ), filtered, and evaporated to give the crude 1,4-products. After purification by4

column chromatography (SiO , hexane:diethyl ether 3:1) the e.e.'s were determined.2

Yields of 6.47-6.49 ca. 50% with e.e.'s of 56%, 62%, and 65%, respectively.

3-(5-Acetoxypentanyl)cyclohexan-1-one (6.47)H NMR * 1.29 (bs, 7H), 1.51-2.06 (m, 7H), 2.01 (s, 3H), 2.13-2.42 (m, 3H), 4.01 (t, J =1

6.6 Hz, 2H). C NMR * 20.71 (q), 24.98 (t), 25.68 (t), 26.01 (t), 28.23 (t), 30.99 (t) ,13

36.15 (t), 38.70 (d), 41.22 (t), 47.89 (t), 64.20 (t), 171.07 (s), 211.88 (s). HRMS calcdfor C H O : 226.157, found 226.157. For e.e. determination, see 6.33 .13 22 3

4,4-Dimethyl-3-(5-acetoxypentanyl)cyclohexan-1-one (6.48)H NMR * 0.98 (s, 3H), 1.03 (s, 3H), 1.03-1.78 (m, 12H), 2.05 (s, 3H), 2.23-2.49 (m, 3H),1

4.08 (t, J = 6.7 Hz, 2H). C NMR * 18.33 (q), 21.15 (q), 25.99 (t), 26.86 (t), 27.05 (t),13

28.43 ( t), 28.67 (q), 32.49 (t), 36.11 (t), 38.57 (d), 40.80 (t), 43.95 (t), 64.21 (t), 170.91(s), 212. 18 (s). HRMS calcd for C H O : 254.188, found 254.188. For e.e .15 26 3

determination, see 6.33 .

1,3-Diphenyl-8-acetoxy-octan-1-one (6.49)H NMR * 0.86-1.63 (m, 8H), 1.96 (s, 3H), 3.01-3.17 (m, 3H), 3.88 (t, J = 6.6 Hz, 2H),1

6.95-7.43 (m, 8H), 7.60-7.79 (m, 2H). C NMR * 20.64 (q), 25.59 (t), 27.14 (t), 28.0813

(t), 33.87 (t), 41.66 (d), 42.38 (t), 64.24 (t), 126.22 (d), 127.70 (d), 128.11 (d), 128.41(d), 128.56 (d) , 132.82 (d), 136.89 (s), 143.03 (s), 198.76 (s). HRMS calcd for C H O :22 26 3

338 .188, found 338.188. E.e. determinated by HPLC analysis; Daicel Chiralcel OD ,5.0% iPrOH in hexane, flow rate 1.0 ml/min, UV detector (236 nm).

Asymmetric amplificationThe enan tiomeric excess of 6.9 was adjusted by mixing appropriate amounts o fen antiomeric pure 6.9 and racemisch 6.9 to give 26.0 mg (0.063 mmol) of scalemi cligand , which was dissolved together with Cu(OTf) (0.03 mmol) in toluene (5 ml) .2

Cha lcone (208 mg, 1.0 mmol) was added and the solution was cooled to -20 EC an ddiethylzinc in toluene (1.1 M, 1.5 equivalent) was added. Stirring was continued at -


141

1. Churchill, M.R.; Kalra, K.L. Inorg. Chem. 1974 , 13, 1899.

2. Floriani, C.; Jacoby, D.; Chiesa-Villa, A.; Guastini, C. Angew. Chem., Int. Ed. Engl. 1989 , 28, 1376.

3. Dr. R. Hulst, unpublished results.

4. a) Schiff, D.E.; Richardson, Jr., J.W.; Jacobson, R.A.; Cowley, A.H., Lasch, J.; Verkade, J.G. Inorg.Chem. 1984 , 23, 3373, and references therein. b) Pastor, S.D.; Hyun, J.L.; Odorisio, P.A.; Rodebaugh,R.K. J. Am. Chem. Soc. 1988 , 110, 6547 and references therein.

5. For an extensive review, see: Asymmetric Synthesis, Chiral Catalysis; Morrison, J.D. Ed.; AcademicPress: New York, 1985 , Vol. 5.

6. a) Tkatchenko, I. In Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F.G.A.; Abel,E.W., Eds.; Pergamon: Oxford, 1982 , Vol. 8, Chapter 50.3. b) Gladiali, S.; Bayón, J.C.; Claver, C .Tetrahedron: Asymmetry 1995 , 6, 1453.

7. Tolman, C.A. Chem. Rev. 1977 , 77, 313.

8. See also: van Rooy, A. Rhodium Catalysed Hydroformylation with Bulky Phosphites as ModifyingLigands, Ph.D. Thesis, University of Amsterdam, 1995 , Chapter 7.

9. 2,2'-Binaphthol (and its derivatives) is one of the most investigated and most widely applied chiralauxiliary in a variety of asymmetric reactions. It is optically stable and it has C -symmetry. See for2

example: a) Stock, H.T. Chiral Thiocrown Ethers, Synthesis and Application in Asymmetric Catalysis,Ph.D. Thesis, University of Groningen, 1994 , Section 2.5 and references therein.For a discussion on C -symmetry, see: b) Whitesell, J.K. Chem. Rev. 1989 , 89, 1581.2

10EC for 16 h and the mixture was poured into 25 ml of aqueous 1 N HCl and extractedwith diethyl ether (3 x 20 ml). The combined organic layers were washed with brine (25ml), dried (MgSO ), filtered and evaporated to give the crude 1,4-product. Afte r4

purificatio n by column chromatography (SiO , hexane:diethyl ether 5:1) the e.e.values2

of 6.5 were determined. E.e. values of 6.9 and 6.5 are shown in Figure 6.4.

Acknowledgements

Mr. A. Meetsma is acknowledged for the determination of the crystal structur edescribed in this Chapter. A. Arnold is thanked for the determination of the nonlineareffect and for the performance of a number of experiments described in this Chapter.Prof. P. Knochel, University of Marburg, is acknowledged, for the opportunity givento perform experimental work in his laboratories. Mr. M. Suijkerbuijk and the NMRservice team (especially Mr. W. Kruizinga and Dr. J. Herrema) are thanked fo rassistanc e with the many e.e. determinations using HPLC and NMR analysis ,respectively.


Chapter 6

142

10. Hulst, R.; de Vries, N.K.; Feringa, B.L. Tetrahedron: Asymmetry 1994 , 5, 699.

11. Greene, N.; Kee, T.P. Synth. Commun. 1993 , 23, 1651.

12. Kindly provided by Dr. C. Leung.

13. Kindly provided by Dr. T. Vries. For the synthesis, see: Vries, T.R. Chiral Cyclic Derivatives of C-2Symmetrical Butanedioic Acids, Ph.D. Thesis, University of Groningen, 1996 .

14. Kindly provided by R. La Crois. Forthcoming Thesis, University of Groningen.

15. For a review, see: a) Narasaka, K. Synthesis 1991 , 1.Two typical examples are already given here:Diels-Alder reactions: b) Ishihara, K.; Yamamoto, H. J. Am. Chem. Soc. 1994 , 116, 1561.Michael additions: c) Cram, D.J.; Sogah, D.Y. J. Chem. Soc., Chem. Commun. 1981 , 625.

16. Cram, D.J.; Helgeson, R.C.; Peacock, S.C.; Kaplan, L.J.; Domeier, L.A.; Moreau, P.; Koga, K.; Mayer,J.M.; Chao, Y.; Siegel, M.G.; Hoffman, D.H.; Sogah, G.D.Y. J.Org. Chem. 1978 , 43, 1930.

17. Lingenfelter, D.S.; Helgeson, R.C.; Cram, D.J. J. Org. Chem. 1981 , 46, 393.

18. Cox, P.J.; Wang, W.; Snieckus, V. Tetrahedron Lett. 1992 , 33, 2253. See also references cited therein.

19. a) Suzuki, A.; Pure Appl. Chem. 1991 , 63, 419. b) Martin, A.R.; Yang, Y. Acta Chem. Scand. 1993 , 47,221. c) A.M. Schoevaars, personal communication. See also reference 18.

20. For a dramatic counterion effect in the chiral Lewis acid catalysed Diels-Alder reaction, see: Evans,D.A.; Murry, J.A.; von Matt, P.; Norcross, R.D.; Miller, S.J. Angew. Chem. Int. Ed. Engl. 1995 , 34, 798.

21. See also a very recent communication of Noyori and co-workers: Kitamura, M.; Miki, T.; Nakano, K.;Noyori, R. Tetrahedron Lett. 1996 , 37, 5141. They observed in the copper (CuCN, CuOTf, CuO- t-Bu, CuCl, CuBr, and CuI) catalysed conjugateadditi on of diethylzinc to 6.2 no synthetically usefull yield of 1,4-product, however, with a smal lamount of an achiral sulfonamide rapid reaction took place (1 h at 0 E C) to afford 6.3 in high yield (>90%, GC analysis).

22. Alexakis, A.; Frutos, J.C.; Mangeney, P. Tetrahedron: Asymmetry 1993 , 4, 2431.

23. Posner, G.; Frye, L.L. Isr. J. Chem. 1984 , 24, 88.


25. Cu salts are used as catalyst in the conjugate addition of Grignard reagents to ",$-unsaturated esters,II

see: a) Sakata, H.; Aoki, Y.; Kuwajima, I. Tetrahedron Lett. 1990 , 31, 1161.However, the exact role of Cu is unclear. In general Cu compounds are reduced with, for example,II II

Grignard reag ents or alkyllithiums to Cu . See: b) House, H.O.; Respess, W.L.; Whitesides, G.M. J. Org.I

Chem. 1966 , 31, 3128. c) House, H.O. Acc. Chem. Res. 1976 , 9, 59. d) See also Section 3.6 for a relatedreduction of nickel .II

26. Dee lman, B.-J.; Bijpost, E.A.; Teuben, J.H. J. Chem. Soc., Chem. Commun. 1995 , 1741. See als oreference 21.

27. Kindly provided by S. Otto. Forthcoming Thesis, University of Groningen.

28. a) Knochel, P.; Singer, R.D. Chem. Rev. 1993 , 93, 2117. b) Langer, F.; Devasagayaraj, A.; Chavant, P.-Y.; Knochel, P. Synlett 1994 , 410.

29. The term sca lemic refers to unequal mixtures of enantiomers. Heathcock, C.H.; Finkelstein, B.L.; Jarvi,E.T.; Radel, P.A.; Hadley, C.R. J. Org. Chem. 1988 , 53, 1922.


143

30. a) Puchot, C.; Samuel, O.; Dunach, E.; Zhao, S.; Agami, C.; Kagan, H.B. J. Am. Chem. Soc. 1986 , 108,2353. b) Noyori, R.; Kitamura, M. Angew. Chem. Int., Ed. Engl. 1991 , 30, 49.See also: c) Wynberg, H.; Feringa, B.L. Tetrahedron, 1976 , 32, 2831.

31. Thiele, K.-H.; Köhlr, J. J. Organomet. Chem. 1968 , 12, 225. See also reference 28a.

32. a) Ullenius, C.; Christenson, B. Pure Appl. Chem. 1988 , 60, 57. b) Krause, N.; Wagner, R.; Gerold, A.J. Am. Chem. Soc. 1994 , 116, 381.

33. For organocopper reagents derived from dialkylzinc reagents, see: Arai, M.; Kawasuji, T.; Nakamura,E. J. Org. Chem. 1993 , 58, 5121, and references therein. See also reference 21.

34. Noyori, R. Asymmetric Catalysis in Organic Chemistry; Wiley: New York, 1994 , Chapter 6.

35. Synthesised by A.M. Schoevaars according to a literature procedure: Coulson, D.R. Inorg. Synth. 1972 ,13, 121.

36. a) Walker, N.; Stuart, D. Acta Cryst. 1983 , A39, 158-166. b) Spek, A.L. HELENA, program for datareduction, Utrecht University, The Netherlands, 1993 .

37. The DIRDIF program system; technical report of the crystallography laboratory of the University ofNijmegen, The Netherlands, 1993 .

38. Xtal3.2 Reference Manual; Hall, S.R.; Flack, H.D.; Stewart, J.M., Eds.; Universities of Wester nAustralia, Geneva and Maryland; Lamb: Perth, 1992 .

39. Spek, A.L. Acta Cryst. 1990 A46, C-34.

40. Meetsma, A. Extended version of the program PLUTO, University of Groningen, The Netherlands,1992 (unpublished).

139

Chapter 7

Asymmetric Catalysis with Phosphorus Amidites

7.1 Introduction

In the previous Chapters we have seen that the enantioselective conjugate addition ofdialkylz inc reagents to several enones is successfully catalysed by complexes derivedfrom a copper salt and a chiral phosphorus amidite. Especially the observed ligand-acceleration in these copper mediated reactions, resulting in a fast and enantioselectiveconjugate addition, has stimulated us to examine this combination in other reactions.Therefore, catalysts derived in situ from copper salts (or in one case a rhodium salt) andchira l phosphorus amidites were used in preliminary experiments of other additio nreactions (mainly carbon-carbon bond forming reactions). As already described i nSection 1.2, t here are several examples known of copper catalysed asymmetric carbon-carbon bond formations for a wide variety of substrates, furnishing products in highyields and with excellent enantioselectivities. Although these catalytic reactions areeligible candidates to examine, we have focussed our attention to addition reactionswhich lack highly enantioselective catalysts with a wide applicability.

7.2 Catalytic Michael addition reaction

The addition of methyl-1-oxo-2-indanecarboxylate to methyl vinyl ketone (MVK)Carbon-carbon bond formation via Michael additions are most frequently performedunde r conditions of base catalysis, however, conjugate addition of 1,3-dicarbony lcompoun ds to enones can also be efficiently catalysed by metal complexes. Al lsuccessful catalytic enantioselective examples are given in Section 2.6. Brunner andHammer were the first to report significant enantioselectivity in a cobalt catalyse dMichael addition of methyl-1-oxo-2-indanecarboxylate ( 7.1) to methyl vinyl ketone(MVK, 7.2 , Eq. 7.1). With (1 S,2S)-(-)-1,2-diphenylethylene diamine as chiral ligand1

the Mic hael product 7.3 was formed with an enantioselectivity of 66%. Later on chiralcopper complexes 7.4a-d were successfully applied in the same reaction with e.e.'s upto 70% (see Section 2.6). Although the enantioselectivity strongly depends on th e2

solvent (low e.e. in toluene, highest values in CCl ) and the chiral catalyst (for example,4

with 7.4b an e.e. of 7% was found), this report prompted us to examine the combinationof chiral phosphorus amidite 7.5b and Cu(OTf) as catalyst in the reaction given in Eq.2

7.1. With 4 mol% of Cu(OTf) and 8.5 mol% of 7.5b the addition of 7.1 to 7.2 occurred2

7.1 7.2

(7.1)O

+

7.5b (8.5 mol%)Cu(OTf)2 (4 mol%)

toluene, CH2Cl2

OMe

O OOMe

O

O O

7.3

N

S Cu

7.5a R = Me7.5b R = i-Pr

R

R''

O

OR' La

OO

7.6

2

NO

O

R

Cu

7.4a R = Et7.4b R = Ph7.4c R = (CH 2)3OH7.4d R = CH2OH

OO

P NR

R

7.7

Chapter 7

140

at am bient temperature in a solvent mixture of toluene and CH Cl in almos t2 2

quantitative yield (91%). In the literature reports the e.e. values were based on the1,2

maximum optical rotation of Michael product 7.3 . However, control experiments by3

Kell er in our research group revealed that these values are not in agreement wit henantiomeric excesses determined by chiral HPLC analysis (Regis; ( R,R)-Whelk-O 1CSP). Unfortunately, the e.e. determination of 7.3 by the latter method showed that no4

enant ioselectivity was induced in the Cu(OTf) / 7.5b catalysed Michael addition (e.e.2

< 5%).

Compared to the enantioselective copper catalysed conjugate addition of dialkylzincreagents to enones ( Chapter 6) a stereogenic carbon center at a different position (at theMichael don or instead of the $-position of the enone) is formed in this reaction .Therefore , we have examined the Michael addition of dibenzyl malonate t ocyclohexenone creating the stereogenic center at the $-position of the enone.

The addition of dibenzyl malonate to cylohexenoneRecently, Shibasaki and co-workers reported a chiral lanthanum complex 7.6 , whichis highly effective as catalyst in enantioselective Michael additions of malonates t ocyclic enones (yield > 90%, e.e. 75-95%). The mode of addition in the preparation of5

the ester enolate 7.6 is crucial for this base catalysis (see also Section 2.6). When weexamined the additio n of dibenzyl malonate to cylohexenone in the presence of 4 mol%of Cu (OTf) and 8.5 mol% of 7.5b we were disappointed that even after 15 days i n2

toluene / THF at -15 EC no trace of the Michael product was detected with TLC .Probably the presence of lantanum is essential for successful catalysis.

R = alkyl, aryl, vinyl, allylY = Cl, Br, OC(O)R, SO2Ph, OR, OP(O)(OR)2

R' Y (7.2)Cu (cat.)I

RM

R'

R'SN2'

R' R SN2

Cu (cat.)7.5a (cat.)

THFMe2Zn+

PhPh Cl (7.3)

7.8 7.9


141

7.3 Copper catalysed S 2' reactionN

IntroductionThe substitution reaction of organometallic reagents (RM = RMgX, R Zn, RZnX, etc.)2

wi th an allylic substrate promoted or catalysed by copper(I) salts or complexes i sreceiving increasi ng attention as a method for regio- and stereoselective carbon-carbonbond formation . Depending on the reaction conditions used, the regioselectivity could6

be directed to afford either the product by S 2 or S 2' reaction (Eq. 7.2).N N

The creati on of a new stereogenic center in the S 2' reaction has led to highl yN

diastereoselecti ve reactions using a chiral allylic substrate, but to our knowledge there7

is only one example known using achiral substrates and a chiral catalyst. The S 2'8N

rea ction of n-BuMgI with two allylic acetates [R'CH=CHCH OAc (R' = PhOCH - or2 2

cycl ohexyl)] catalysed by the chiral arenethiolatocopper(I) complex 7.7 proceede dwith an enantioselectivity up to 42%. 9

The addition of dimethylzinc to cinnamyl chlorideSince it is k nown that organozinc reagents (RZnX and R Zn) does not react with allylic2

ac etates in the presence of copper(I) complexes, we have examined the coppe r10

catalyse d addition of dimethylzinc to cinnamyl chloride ( 7.8) in the presence of chiralphosphoru s amidite 7.5a (Eq. 7.3). The S 2' reaction catalysed by a mixture of CuCN11

N

(20 mol%), LiCl (20 mol%), and 7.5a (20 mol%) in THF as well as by Cu(OTf) (102

mol%) / 7.5a (20 mol%) in THF proceeded smoothly furnishing the desired product 7.9in good yield (ca. 80%).

Chapter 7

142

Unfortu nately, the e.e. determination by chiral GC analysis (Lipodex C column )revealed no enantioselectivity for both cases. When the reactions were performed withdiethy lzinc instead of dimethylzinc the reaction given in Eq. 7.3 gave th eco rresponding product as well, however, the e.e. could not be determined using th emethod mentioned above (no separation of both enantiomers).Proba bly the formation of a key-intermediate in which the allylic substrate anchors ina bidentate fashion [ B-complexation of the C=C double bond to copper an dcoordin ation of the carbonyl-oxygen atom of the acetate function to M (for M, see Eq.7.2)] to the chiral copper complex, as proposed in the literature, is not possible with9

cinnam yl chloride resulting in a non-enantioselective reaction. It should be noted thatallylic phos phates undergo S 2' reaction with dialkylzinc reagents and this conversionN

deserves to be examin ed with our chiral catalytic system. Moreover, other combinationsof organometallic reagent and substrates in the presence of copper / 7.5 are possiblymore successful.

7.4 Catalytic asymmetric addition of organometallic reagents to imines

IntroductionCompared to the addition to carbonyl compounds, examples of asymmetric addition ofcarbo n nucleophiles to imines have been reported scarcely. Two successfu lcon tributions have been emphasised in Section 1.2: the addition of organolithiu mreagent s to N-arylimines (e.e.'s up to 91%) catalysed by a C -symmetric chira l2

bis(oxazoline) ligand and an effective Strecker synthesis employing a chiral cyclic12

dipeptide as catalyst with exceptionally high e.e.'s (e.e.'s up to 99%). 13

Imines are unreactive to diethylzinc even in the presence of stoichiometric amounts ofamino alcohol promoters. Enantioselective alkylations with dialkylzinc reagents arereported to act ivated imines ( N-diphenylphoshinoylimines, 'masked'- N-acylimines, andnitrones) in the presence of (sub)stoichiometric amounts of chiral amino alcohols .14

Recently, the use of a catalytic amount of chiral auxiliary (L*MgBr, L* = (2 S,3R)-4-dimethylamino-1,2-diphenyl-3-methyl-2-butoxide) was successful in th eenantioselective addition of dialkylzinc reagents to nitrones (e.e. 56-78%), however,the presence of Ph COMgBr (0.3 equivalent) is essential, probably to regenerate the3

chiral auxiliary. 15

The addition of diethylzinc to 'activated' isoquinolineIn the research g roup of Prof. Kellogg the reaction of diethylzinc to isoquinoline ( 7.10 )has been examined by Naasz. In the presence of stoichiometric amounts of (chiral)16

amino alcohols no reaction occured between diethylzinc and isoquinoline either a t

N +Cl OPh

O

N OPh

O

7.10 7.11

Cu(OTf)2 (10 mol%)7.5b (21 mol%)

Et2Zn, toluene(7.4)


143

elevated temperatures or by activation with Lewis acids [BF , Ti(O i-Pr) ]. However ,3 4

reaction of isoquinoline with phenyl chloroformate resulted in a activated iminium ionwhich was success fully treated with diethylzinc. Reactions at -70 EC in toluene gave theethyla ted product 7.11 , but with no enantioselectivity, even in the presence o fstoichiometric amounts of several chiral amino alcohols. 16

When we added a clear solution of Cu(OTf) (10 mol%), phosphorus amidite 7.5b (212

mol%), and Et Zn (1 mmol) in 5 ml of toluene to a solution of isoquinoline and phenyl2

chloroformate in 10 ml of toluene at -50 EC, the ethylated product 7.11 was isolate dafter 16 h in 84% yield (Eq. 7.4). Unfortunately, the e.e. determination by chiral HPLCanalysis (Daicel; Chiralpak AD) revealed no enantioselectivity. This lack o fenantioselectivity is probably due to the high rate of the uncatalysed reaction underthese reaction conditions (without chiral promoter the reaction proceeds at the samerate). Less reactive substrates, for example nitrones, or even non-activated imines in16

the presence of a more reactive alkylating reagent seem to be feasible options fo renantioselective additions to imines.

7.5 Addition of diethylzinc to tropone

2,4,6-C ycloheptatrien-1-one (tropone, 7.12 ) can be functionalised in a 1,8-fashion viathe ad dition of nucleophiles such as enolates and Grignard reagents to give th ecorrespondin g 2-substituted dihydrotropone. We were interested whether diethylzinc17

can be added (enantioselective) to tropone in a 1,8-fashion as well. Therefore, to aso lution of Cu(OTf) (3 mol%) and phosphorus amidite 7.5b (6 mol%) in toluene ,2

tropo ne and Et Zn were added successively at ambient temperature (Eq. 7.5). Th e2

colour of the mixture changed from yellow to dark red and after 48 h the mixture wasworked up (TLC revealed the disappearance of the starting material). However, amixture of at least four products was obtained and we were not able to isolate one ofthem fr om this mixture. The low reactivity of diethylzinc (or the transmetalate dreagent) seems to be the limiting factor for selective addition of diethylzinc to tropone.

7.12

O

Cu(OTf)2 / 7.5b (7.5)mixture of unidentified materials+ Et2Zn

7.13

O

O

Cu(OAc)2, L*,copper bronze

t-BuO3CPhEtCO2H

(7.6)

Chapter 7

144

7.6 Copper catalysed enantioselective allylic oxidation

The functionalisation of alkenes and alkanes by catalytic enantioselective oxidationshas been investigated to a wide extent. Besides the highly enantioselectiv e18

ep oxidation of allylic alcohols and unfunctionalised alkenes, and th e18a 18b

dihydroxylation of a wide range of alkenes, the asymmetric allylic oxidation is an18c

interesting alternative for direct functionalisation of alkenes. In our research group19

the copper cata lysed (0.5 mol% Cu(OAc) and 5 mol% copper bronze) allylic oxidation2

of 2-cyclo hexene in the presence of propionic acid and peresters to yield th eallylpropionate has been examined (Eq. 7.6). With ( S)-proline as chiral ligand (L*, 3mol%), 2-cycloh exenyl propionate ( 7.13 ) was obtained in ca. 60% yield (relative to theperesters) with e.e.'s up to 61%. 19

When this copper catalysed allylic oxidation was performed in the presence o fphosphorus a midite 7.5a (3 mol%), the conversion to product 7.13 (70% relative to theoxidant) proceeded with a selectivity and at a rate comparable with the values foundwith (S)-proline as ligand. An e.e. of 19% was determined by chiral GC analysis.19 20

Both the conversion to the product and the observed enantioselectivity are a nin dication of a role of the chiral ligand in this catalytic oxidation. Investigations t ooptimise the conditions for this specific chiral ligand still has to be performed.

7.7 Rhodium catalysed enantioselective hydroformylation

The hydroformylation reaction of alkenes has enormous potential for the synthesis ofoptically active aldehydes and especially in the last three years significant advances

7.157.14

(7.7)+Rh(acac)(CO)2 / 7.5a

CO / H2

CHOCHO


145

have bee n achieved using chiral ligands containing phosphorus donors. Achira l21

phospho rus amidites have been applied as ligands in the rhodium catalyse dhydrofo rmylation of 1-octene and styrene. In cooperation with the research group of22

Prof. van Leeuwen we have examined the chiral phosphorus amidite 7.5a as ligand inthe rhodium catalysed hydroformylation of styrene (Eq. 7.7). 23

The catalyst was formed with Rh(acac)(CO) (0.1 mol%) and amidite 7.5a with 20 bar2

CO/H at 50 EC in toluene for 2 h and successively treated with styrene under the same2

conditio ns. With a ligand to rhodium ratio of 10 the conversion after 4 h was 33 %(branched/linear ratio of 7) which is comparable with the values observed for achiralphospho rus amidites. Unfortunately, the e.e. determination of the correspondin g22

alcohol of 7.14 by chiral GC revealed no enantioselectivity. With a ligand to rhodiumratio of 50 the catalyst is hardly active (4 % conversion after 4 h).In practically all chiral catalysts for enantioselective hydroformylation reactions theligands are bidentate phosphorus compounds, so probably better results will b eobtained with a bidentate phosphorus amidite.

7.8 Concluding remarks

The prelim inary experiments to develop other enantioselective addition reactions witha catalyst derived from a chiral phosphorus amidite revealed that these attempts havenot been very successful. Only in the case of the allylic oxidation of 2-cyclohexene agood co nversion to the product was accompanied with a significant enantioselectivity(19%). Although the product obtained via this reaction can be easily converted t oin teresting compounds, experiments to enhance the enantioselectivity seems to b eindistinct.In the Michael addition of the indanecarboxylate 7.1 to MVK the lack o fenantiose lectivity can probably be explained by the mechanism of the reaction. In thiscase the 1,3- dicarbonyl compound is activated by the copper complex prior to reactionfurnishing a product with a stereogenic center at a different position (at Michael donorinstead of enone) as compared with the products obtained with the conjugate addition

Chapter 7

146

of dialkylzinc reagents to enones (described in the previous Chapters). Other ligandsseems to be required for these reactions.The addition of diethylzinc to allylic substrates, imines, and tropone described in thisChapter probably all proceed via the same organometallic reagent as proposed for thecopper catalysed conjugate addition of dialkylzinc reagents to enones (Chapter 5 and6). Unfortunately, with tropone or allylic acetate this alkyl-transferring reagent is notsufficiently reacti ve compared to the Grignard reagents, resulting in unwanted productsor no addition reaction, respectively. The addition to allylic chloride 7.8 resulted in aS 2' reac tion with no enantioselectivity, probably due to the lack of a highly regulatedN

transiti on state - bidentate anchoring of the substrate to the chiral copper comple xseems to be required. The lack of enantioselectivity in the addition of diethylzinc to anactivat ed imine (iminium ion) is probably due to competing uncatalysed addition. Thesearch fo r the delicate balance between unreactive and slightly reactive substrate sseems to be the determinant factor for this addition reaction.Although the rhodium catalysed hydroformylation proceeded in the presence of 7.5awith a rat e comparable as found for achiral phosphorus amidites, no enantioselectivitywa s observed. The use of bidentate phosphorus amidite ligands seem to be feasibl eoptions for this reaction.


All reactions in this Chapter were performed under argon using flame-dried standardSchlenk equipment, u nless stated otherwise. For more general remarks, see Sections 3.8and 6.8.

MaterialsThe follow ing compounds were commercially available and used without purification:MVK (Aldrich), dibenzyl malonate (Aldrich), cinnamyl chloride ( 7.8 ; Aldrich) ,isoquinoline ( 7.10; Aldrich), phenyl chloroformate (Aldrich), tropone ( 7.12 ; Lancaster),2-cyclohexene (Aldrich), Cu(OAc) CH O (Aldrich), t-butylperoxy benzoate (Aldrich),2 2

and propionic acid (Aldrich).Me thyl-1-oxo-2-indanecarboxylate (7.1) was prepared by E. Keller according to aliterature procedure. For all other materials, see Sections 3.8, 5.6, and 6.8.24

Addition of methyl-1-oxo-2-indanecarboxylate to MVK (preparation of 7.3)To a cooled (-15 EC) solution of Cu(OTf) (15 mg, 0.04 mmol) and 7.5b (36 mg, 0.0852

mmol) in toluene (4 ml) and CH Cl (5 ml) methyl-1-oxo-2-indanecarboxylate (0.19 g,2 2


147

1.0 mmol ) and MVK (190 mg, 2.7 mmol) were added. The mixture was stirred a tambient temperatu re for 3 days. The conversion of the indanone was confirmed by TLCana lysis and the mixture was evaporated and purified by column chromatograph y(SiO , hexane:EtOAc 9:1) to afford the pure Michael product in 91% yield. The H and2

1

C NMR sp ectra were in agreement with those reported in the literature. The e.e. was13 1,2

determined by HPL C analysis; Regis, ( R,R)-Whelk-O 1 CSP, 20% EtOH in hexane, flowrate 1.0 ml/min (see also reference 4).

Addition of dimethylzinc to cinnamyl chloride (preparation of 7.9)This procedure is typical for the S 2' reactions described in this Chapter. To a cooledN

(-25EC) solution of Cu(OTf) (36 mg, 0.10 mmol) and 7.5a (80 mg, 0.20 mmol) in THF2

(3 ml) Me Zn (1.0 ml, 2M in toluene, 2 mmol) was added. The resulting clear yellow2

solution was cooled to -60 EC and cinnamyl chloride (0.16 g, 1.05 mmol) was added.The solut ion was stirred for 1 h at -60 EC and 15 h at ambient temperature. The mixturewas poured in to 25 ml of aqueous 1 N HCl and extracted with diethyl ether (2 x 25 ml).The co mbined organic layers were dried (MgSO ), filtered, and evaporated to give the4

crud e product 7.9 . The H and C NMR spectra were in agreement with literatur e1 13

reports. The e.e. was determined by chiral GC analysis (Lipodex C column).6,11

Addition of diethylzinc to 'activated' isoquinoline (preparation of 7.11)A pref ormed clear orange solution of Cu(OTf) (36 mg, 0.10 mmol), 7.5b (88 mg, 0.212

mmol), and Et Zn (1.1 ml, 1.1M in toluene, 1.2 mmol) in toluene (5 ml) (prepared at -2

10EC and stirred for 15 min at ambient temperature) was added by syringe in 5 min toa cooled (-5 0EC) mixture of isoquinoline (118 µl, 1.0 mmol) and phenyl chloroformate(125 µl, 1.0 mmol) in toluene (10 ml). The mixture was stirred for 1 h at -50 EC and 15h at ambient temper ature. The obtained clear solution was poured into 25 ml of aqueoussaturated NH Cl and extracted with diethyl ether (2 x 25 ml). The combined organic4

layers wer e washed with aqueous 1 N NaOH (25 ml) and brine (25 ml), dried (MgSO ),4

filtered, and evapora ted to give the crude product. This material was purified by columnchromatography (SiO , CH Cl :hexane 2:1) to afford pure 7.11 (yield 84%) as a white2 2 2

solid. H and C NMR were in agreement with those reported by R. Naasz. The e.e.1 13 16

was determined by chiral HPLC analysis (Daicel; Chiralpak AD, 5.0% iPrOH in hexane,flow rate 0.5 ml/min).

Allylic oxidation of 2-cyclohexene (preparation of 7.13)This experiment was performed by C. Zondervan, University of Groningen. Ligand 7.5a(0.12 g, 0.33 mmol), Cu(OAc) CH O (9.0 mg, 0.05 mmol), and copper bronze (35 mg,2 2

0.55 mmol) we re suspended in acetonitrile (3 ml), 2-cyclohexene (3 ml), and propionic

Chapter 7

148

1. Brunner, H.; Hammer, B. Angew. Chem., Int. Ed. Engl. 1984 , 23, 312.

2. a) Desimoni, G.; Quadrelli, P.; Righetti, P.P. Tetrahedron 1990 , 46, 2927. b) Desimoni, G.; Dusi, G.;Faita, G.; Quadrelli, P.; Righetti, P. Tetrahedron 1995 , 51, 4131.

3. Hermann, K.; Wynberg, H. J. Org. Chem. 1979 , 44, 2238.

4. E. Keller, unpublished results. See forthcoming thesis, University of Groningen.

5. Sasai, H.; Arai, T.; Shibasaki, M. J. Am. Chem. Soc. 1994 , 116, 1571.

6. a) Magid, R.M. Tetrahedron 1980 , 36, 1901. b) Bäckvall, J.-E.; Sellén, M.; Grant, B. J. Am. Chem. Soc.1990 , 112, 6615. c) Goering, H.L.; Underiner, T.L. J. Org. Chem. 1991 , 56, 2563. d) Arai, M. ;Nakamura, E.; Lipshutz, B. J. Org. Chem. 1991 , 56, 5489. e) Arai, M.; Kawasuji, T.; Nakamura, E. J.Org. Chem. 1993 , 58, 5121. f) Yanagisawa, A.; Nomura, N.; Yamamoto, H. Tetrahedron 1994 , 50,6017. g) Persson, E.S.M.; van Klaveren, M.; Grove, D.M.; Bäckvall, J.-E.; van Koten, G. Chem. Eur. J.

acid under a nitrogen atmosphere. After stirring for 25 min at ambient temperature agreen suspension was obtained and t-butylperoxy benzoate (1.0 ml, 5.0 mmol) wa sadded. The mixture was stirred for 5 days at ambient temperature and poured into 25ml o f aqueous 2 N HCl and extracted with diethyl ether (2 x 25 ml). The combine dorganic laye rs were washed with saturated aqueous NaHCO (25 ml) and brine (25 ml),3

dried (NaS O ), filtered, and evaporated to give the crude product 7.13 . The conversion4

and e.e. were determined by chiral GC analysis (see text for data)

Rhodium catalysed hydroformylation of styreneThis exp eriment was performed by S. Deerenberg, University of Amsterdam. In a nevacuated autoclave the catalyst was formed with Rh(acac)(CO) (0.02 mmol) an d2

amidit e 7.5a (0.20 mmol) with 20 bar CO/H at 50 EC in toluene (5 ml) for 2 h. Th e2

resulting mixture was treated with a solution of styrene (2.3 ml, 20 mmol) in toluene(6.7 ml) u nder the same conditions and after 4 h the conversion was determined by GCanalysis (3 3%) (branched/linear ratio of 7). The aldehydes 7.14 and 7.15 were reducedto the corresponding alcohols, prior to e.e. determination (chiral GC analysis).

AcknowledgementsC. Zondervan, University of Groningen, and S. Deerenberg, University of Amsterdamare great fully acknowledged for the performance of the allylic oxidation and th ehydroformyl ation, respectively. E. Keller is thanked for the synthesis of compound 7.1and for discussions concerning the Michael addition reactions and the addition reactionto tropone. Furthermore, H. van der Worp and R. Naasz are thanked for discussionsconcerning S 2' reactions and diethylzinc additions to imines, respectively.N



149

1995 , 1, 351.

7. Denmark, S.E.; Marble, L.K. J. Org. Chem. 1990 , 55, 1984 and references therein. See also references6d and 6e .

8. The palladium catalysed enantioselective allylic alkylations of 1,3-disubstituted-2-propenyl acetates,des cribed in Section 1.2, are not included, since they proceed via a symmetrical B-allyl palladiu mcomplex.

9. a) van Klaver en, M.; Persson, E.S.M.; del Villar, A.; Grove, D.M.; Bäckvall, J.-E., van Koten, G .Tetrahedron Lett. 1995 , 36, 3059. b) van Klaveren, M. Arenethiolatocopper(I) Complexes asHomogeneous Catalysts in Organic Synthesis, Ph.D. Thesis, University of Utrecht, 1996 .

10. Sekiya, K. ; Nakamura, E. Tetrahedron Lett. 1988 , 29, 5155. See also reference 6e. Our ow ninvestigati ons on the reactions with diethylzinc and cinnamyl acetate in the presence of catalyti camounts of CuOTf and 7.5a were in agreement with these reports. Only starting material was isolatedafter 16 h at ambient temperature.

11. H. van der Wor p is thanked for discussions concerning the S 2' reaction. See forthcoming Ph.D. Thesis,N

University of Groningen.

12. Denmark, S.E.; Nakajima, N.; Nicaise, O. J.-C. J. Am. Chem. Soc. 1994 , 116, 8797. See also: Inoue, I.;Shindo, M.; Koga, K.; Tomioka, K. Tetrahedron 1994 , 50, 4429.

13. Iyer, M.S.; Gigstad, K.M.; Namdev, N.D.; Lipton, M. J. Am. Chem. Soc. 1996 , 118, 4910.

14. a) Soai, K.; Hatanaka, T.; Miyazawa, T. J. Chem. Soc., Chem. Commun. 1992 , 1097. b) Katritzky, A.R.;Harris, P.A. Tetrahedron: Asymmetry 1992 , 3, 437. c) Ukaji, Y.; Hatanaka, T.; Ahmed, A.; Inomata, K.Chem. Lett. 1993 , 1313.

15. Ukaji, Y.; Kenmoku, Y.; Inomata, K. Tetrahedron: Asymmetry 1996 , 7, 53.

16. R. Naasz, unpublished results. See forthcoming research report, University of Groningen.

17. Rigby, J.H.; Senanayake, C.H.; Rege, S. J. Org. Chem. 1988 , 53, 4596 and references therein.

18. Chapter 4, in Catalytic Asymmetric Synthesis; Ojima, I., Ed.; VCH Publishers: New York, 1993 . a) Section 4.1; Johnson, R.A.; Sharpless, K.B. b) Section 4.2; Jacobsen, E.N. c) Section 4.4; Johnson,R.A.; Sharpless, K.B.

19. a) Rispens, M.T.; Zondervan, C.; Feringa, B.L. Tetrahedron: Asymmetry 1995 , 6, 661 and referencestherein. b) Zondervan, C.; Feringa, B.L. Tetrahedron: Asymmetry 1996 , 7, 1895.

20. Experiment performed by C. Zondervan, see forthcoming Ph.D. Thesis, University of Groningen.

21. Review: Gladiali, S.; Bayón, J.C.; Claver,C. Tetrahedron: Asymmetry 1995 , 6, 1453.

22. van Rooy, A. Rhodium Catalysed Hydroformylation with Bulky Phosphites as Modifying Ligands, Ph.D.Thesis, University of Amsterdam, 1995 , Chapter 7.

23. Experiments performed by S. Deerenberg, University of Amsterdam.

24. House, H.O.; Hudson, C.B. J. Org. Chem. 1970 , 35, 647.

1

Samenvatting

Algemene achtergrond

Vaak wordt een chemische verbinding geassocieerd met vervuiling of nog erger eenmilieuramp , maar dit beeld is in bijna alle gevallen niet terecht. Helaas zijn er ee naantal gebeurt enissen te noemen waarbij chemische verbindingen ongewenste effectenveroorzaken. Doorgaans spelen chemische verbindingen echter, vaak onbewust, eenbelangrijke rol in het welzijn van de mens. Eenvoudige pijnstillers als aspirine e nparacetamol, meer geavanceerde kanker- en aidsremmende medicijnen, maar oo kallerlei verven, voedingsadditieven, gewasbeschermingsmiddelen, supersterke vezelsen andere kunststoffen zijn alle chemische verbindingen.Een chemische v erbinding bestaat uit allemaal steeds dezelfde hele kleine bouwstenen,moleculen gehet en. Deze moleculen zijn weer opgebouwd uit nog kleinere bouwstenenen die worden atomen (elementen) genoemd. Er zijn meer dan honderd verschillendeelementen maar organisch chemische moleculen zijn meestal opgebouwd uit slechtseen tiental verschillende elementen. Het belangrijkste element van een organisc hmolecul e is koolstof (C), te vergelijken met het bottenskelet van een mens .Ko olstofketens vormen de ruggegraat van organische moleculen waaraan allerle iandere elementen o.a. waterstof (H), - te vergelijken met de spieren van het menselijklichaam - stikstof (N), zuurstof (O), zwavel (S), fosfor (P), etc. - de organen van eenorganisch molecule - gebonden zijn.In de or ganische chemie zijn koolstof-koolstof bindende reakties essentieel, zoiets alsde groei van het menselijk lichaam of de genezing van een botbreuk. Eén bepaald ekools tof-koolstof bindende reaktie is het onderwerp van het onderzoek beschreven indit proefsc hrift. Voordat die reaktie ter sprake gebracht wordt moet nog een belangrijkaspect aangaande dit onderzoek nader belicht worden.

ChiraliteitOrganische molecul en zijn vaak chiraal. Chiraliteit heeft te maken met de vorm van eenbepaald molecule of object. Wanneer iets niet tot dekking te brengen valt met zij nsp iegelbeeldvorm is zo'n molecule of object chiraal. Een voorbeeld van een chiraa lobject zijn o nze handen. Onze handen lijken heel veel op elkaar maar ze zijn nie tprecies hetzelfde, ze zijn spiegelbeelden van elkaar. Een ander voorbeeld van ee nchiraal object staat weergegeven in figuur 1, beide yin & yang symbolen zij nspiegelbeelden van elkaar.

Moleculen die spiegelbeelden van elkaar zijn noemen we enantiomeren. De oorzaakvan deze spiegelbeeldrelatie in chirale moleculen is meestal de aanwezigheid van eenasymmetrisch koolstofatoom. Een koolstofatoom is asymmetrisch als er vie rverschillende groepen of atomen aan gebonden zijn (figuur 1). Enantiomeren van eenchiraal molecule verschillen in één zeer belangrijke eigenschap: ze hebben ee nverschillende interactie met andere chirale moleculen. Dit is te vergelijken met d einteractie van een hand met een linker- en rechterhandschoen (zijn ook chiraal) of vaneen voet met een paar schoenen. De hand of voet past slechts in één handschoen ofschoen.

O

d +

d -

d -

d +

+ Znd - d -

d +

chirale katalysator O

*a

b

chiraal produkt

Samenvatting

3

positieve lading krijgt. Met andere woorden, organische verbindingen reageren me telkaar tot een a ndere verbinding als er een negatieve lading gekoppeld kan worden aaneen positieve lading. Voor de reaktie (de zogenoemde 1,4-additie of geconjugeerdeadditie) die in dit proefschrift uitgebreid bestudeerd is, is de ladingsverdeling van debeide uitgangsstoffen weergegeven (zie figuur 2). Het is belangrijk om te weten datkoolst ofketens van organische verbindingen worden weergegeven met een zigzag -streep, waarbij elk hoekpunt één koolstofatoom voorstelt. Elk koolstofatoom gaat vierbindingen aan m et buuratomen en als er maar 1,2 of 3 bindingen staan aangegeven zijnde resterende groepen waterstofatomen. Ter verduidelijking worden d ewaters tofatomen vaak weggelaten (ter vergelijking: in een schematische tekening vaneen mens wordt bijvoorbeeld een been ook getekend als een streep).

Figuur 2 De geconjugeerde additie van diethylzink aan chalcon (het substraat)gekatalyseerd door een chiraal metaalcomplex. (Ladingsverdeling in hetsubstraat en het reagens zijn weergegeven.)

De lad ingsverdeling in de ene uitgangsstof (het substraat) is het resultaat van d eaanwezigheid van het zuurstof atoom (O). Deze is gedeeltelijk negatief geladen (d -)waardoor het naaste koolstofatoom juist gedeeltelijk positief geladen is (d +). Met eendubbele koolstof-koolstof binding direct naast deze zogeheten carbonyl groep (C=O)wordt de ladingsverdeling uitgebreid naar de koolstofatomen van de C=C binding (ziefiguu r 2, dit heet geconjugeerd). De andere uitgangsstof (het reagens), waarin d eladingsverdeling het resultaat is van de aanwezigheid van een metaalatoom (Zn), kandaardoor me t zijn negatief geladen koolstofatoom op twee verschillende manieren methet substr aat reageren (via pijl a en via pijl b). Dit zou resulteren in twee verschillendeprod ukten maar door de chirale katalysator zo te kiezen wordt in dit geval allee nreaktiepad b gevolgd met als resultaat het produkt weergegeven in figuur 2 (het 1,4-produkt). Dit produkt heeft een koolstofatoom met vier verschillende groepen waardoorhet chiraal is (wordt vaak weergegeven met een *), en er zijn dus ook twe eenant iomeren van het produkt mogelijk. Door nu de juiste chirale katalysator t egebruiken zou deze reaktie enantiospecifiek moeten kunnen verlopen, d.w.z. er wordt

chirale katalysatorEt2Zn+ (1)

Ph Ph

O

Ph Ph

O

*

Samenvatting

4

slechts één enantiomeer van het produkt gevormd. Dit is voor deze reaktie no ghypothetisch en tot nu toe zijn er alleen maar voorbeelden bekend van een katalytischereaktie met een bepaalde voorkeur voor één enantiomeer (enantioselectief) ,weergegeve n met de afkorting e.e.(enantiomeric excess). De e.e. van een verbinding iseen mate va n enantiomere zuiverheid. Een e.e. van 100% wil zeggen dat de verbindingenantiomeer zuiver is en een e.e. van 0% betekent dat van beide enantiomeren evenveelaanwezig is (een racemisch mengsel).Naa st de ontwikkeling van nieuwe katalysatoren voor deze C-C bindingvormend ereaktie zijn we er in geslaagd hoge selectiviteiten voor één van de twee enantiomerente bewerkstelligen.

Samenvatting van het proefschrift

Het werk beschreven in dit proefschrift is vooral gericht geweest op de ontwikkelingvan allerlei chirale katalysatoren, opgebouwd uit een metaal atoom (voornamelij kni kkel of koper) en een chiraal ligand, voor een betere controle op d eenantioselectiviteit voor de reaktie weergegeven in figuur 2.Hoo fdstuk 1 geeft een korte introductie over routes naar enantiomeer zuiver everbindingen en een bondig overzicht van allerlei andere succesvol gekatalyseerd eenan tioselectieve koolstof-koolstof koppelingsreakties. In hoofdstuk 2 is ee nuitgebreid overzicht gegeven van alle eerder beschreven pogingen een gekatalyseerdeenantioselectieve 1,4-additie te verkrijgen. De beste chirale katalysatoren zijn echteralleen maar succesvol (e.e.'s tot 90%) gebleken voor één bepaald substraat. D ebelangri jkste uitdaging is dan ook een algemene katalysator voor de 1,4-additie t eontwik kelen met een goede activiteit en hoge enantioselectiviteit voor meerder esubstraten en reagentia.De belangrijkste chirale katalysatoren en de resultaten van de enantioselectieve 1,4-additi es met het cylische en acyclische modelsubstraat, die beschreven staan in di tproefschrift, zijn weergegeven in de grafische samenvatting.De resultaten beschreven in hoofdstuk 3 laten zien dat de 1,4-additiereaktie va ndiethylzink aan chalcon (acyclisch ",$-onverzadigd keton) gekatalyseerd door een insitu gevormd chiraal nikkelcomplex resulteert in een hoge opbrengst van het 1,4 -produkt met e.e.'s tot 85% (zie vergelijking 1 en grafische samenvatting).

Samenvatting

5

Als chiraa l ligand werden allerlei verschillende $-aminoalcoholen, gesynthetiseerd uithet in de natuur voorkomende (+)-kamfer, gebruikt. Een gedetailleerde studie naar deinvloe d van de struktuur van het chiraal ligand en de reaktiecondities op d eenan tioselectiviteit heeft tot de volgende conclusies geleid: (1) Alleen een cis-configuratie van een tertiair amine en de alkoholfunktionaliteit geeft het 1,4-produktin hoge enantioselectiviteit. (2) Twee equivalenten van het chirale ligand t.o.v. nikkelis al voldoende voor het verkrijgen van een hoge selectiviteit en het toevoegen vanachirale amines heeft geen effect. (3) Andere dialkylzink-

hoofdstuk 3

hoofdstuk 6

hoofdstuk 5

hoofdstuk 4

Ph Ph

O

*

O

*

E.e.'s van de 1,4-produkten

Chirale katalysator(in situ gemaakt)

NHO

NOH

RN

R

OH

-

-

0%

+ Ni(acac) 2

+ Ni(acac) 2

+ Ni(acac) 2

< 5%

75 - 85% (S)

35 - 85% (R)

2( )HO

N

HON + Ni(acac) 2 69% (S) 0%

62% (R)11% (S)+ CuOTfN

S

OPh

N

63% (S)90% (R)+ Cu(OTf)2OO

P N

Samenvatting

6

Grafische samenvatting

(2)

O O

*

chirale katalysatorEt2Zn+

Samenvatting

7

reagentia geven nagenoeg dezelfde goede enantioselectiviteit maar variatie van he tsubst raat, daarentegen resulteert, voor bijvoorbeeld cyclohexenon, in 0% e.e. (zi everge lijking 2 en grafische samenvatting). (4) De katalysator verliest zijn selectiviteitnaar verloop va n tijd en (5) er is een niet-lineair verband aangetoond tussen de e.e. vanhet chirale ligand en de e.e. van het 1,4-produkt, wat een indicatie is voor d eaanwez igheid van twee chirale liganden in het voor de enantioselectiev ealkyloverdracht verantwoordelijke intermediair.Het onderzoek beschr even in hoofdstuk 4 behelst voornamelijk de synthese van nieuwetri- en tetradentaat liganden, wederom uit (+)-kamfer, en de toepassing van dez eliganden in de 1, 4-additiereaktie (zie vergelijkingen 1 en 2). Alhoewel het uiteindelijkedoel om met deze multidentaat liganden een enantioselectieve katalysator te creëerenvoor meerdere typen substraten niet slaagde, is er wel aangetoond dat met dez eliganden een go ede e.e. is te behalen voor het 1,4-produkt van het acyclische substraat.Helaas waren we met de aminoalcohol liganden, beschreven in hoofdstuk 3 en 4, nietin staat om eenduidige nikkelcomplexen te isoleren.In hoof dstuk 5 wordt de invloed van de variaties van het metaalzout, chiraal ligand enreagens op de regio- en enantioselectiviteit van de 1,4-additiereakties beschreven .Additie van diethylzink aan chalcon gekatalyseerd door de chirale aminoalcoholen,beschreven in ho ofdstuk 3, en andere metaalzouten [bijv. Co(acac) en CuBr] resulteert2

in een lagere regioselectiviteit [reaktiepad a (figuur 2) en reduktiereakties vinden ookpla ats], maar nagenoeg dezelfde enantioselectiviteit [t.o.v. Ni(acac) en chiraa l2

aminoalcohol]. Wanneer echter andere chirale liganden worden gebruikt in de koper-gekatalyseerde 1,4-additiereaktie van diethylzink aan cyclohexenon en chalcon is deregioselecti viteit tot het 1,4-produkt zeer hoog (3 uur, > 90%) en zijn de waargenomenenantioselectivit eiten verrassend. Met bijvoorbeeld een chiraal zwavelligand en CuOTf(zie grafisch e samenvatting) verloopt de additie aan chalcon met slechts 11% e.e. maaraan cyclohexenon daarentegen met 62% e.e.

Verander ing van het gebruikte reagens tot een Grignard reagens of ee ntrialkyla luminiumreagens heeft helaas niet geleid tot een 1,4-additie met een hog eenantio selectiviteit. Wel werd er vooral in het geval van een koper-gekatalyseerd etri methylaluminium additie aan chalcon een zeer hoge regioselectiviteit (> 95% )

Samenvatting

8

gevonden maar de enantioselectiviteit bleef steken bij 30% (met een van (+)-kamferafgeleid $-aminoalcohol als chiraal ligand).In hoofdstuk 5 is verder een in situ gevormd complex van een fosforligand, van hettype zoals weergegeven in de grafische samenvatting, samen met CuI gebruikt al schirale katalysator in de reakties geillustreerd in vergelijkingen 1 en 2. Hiermee werdvoor de eerste maal een redelijke enantioselectiviteit verkregen voor zowel de additieaan een cyclisch als aan een acyclisch ",$-onverzadigd keton (35% en 47 %respectievelijk).In hoofdstuk 6 is de struktuur van de nieuwe fosforliganden, met behulp van d egegevens uit de krist alstruktuur van een CuI / ligand complex, geoptimaliseerd. Daartoewerden er een tiental nieuwe fosforliganden, allen behorende tot de nieuwe klasse vanfosforamidaten, gesynthetiseerd. Het chirale fosforamidaat, weergegeven in d egrafisc he samenvatting, bleek voor zowel het cylische als het acyclisch emodels ubstraat de hoogste e.e. te geven in de koper-gekatalyseerde 1,4-additiereaktievan diethylzink. Variatie van de chirale fosforamidaten, reaktiecondities en d euitgangsstoffen heeft geleid tot de volgende conclusies: (1) Ruimtelijk grote groepenop de aminefunctionaliteit van het fosforamidaat leidt tot de hoogste e.e. waarden. (2)Een zeer snelle katalyse en hoge regio- en enantioselectiviteit wordt verkregen me tCu(OTf ) . (3) Hoge regio- en enantioselectiviteiten zijn mogelijk voor allerle i2

vers chillende typen ",$-onverzadigde ketonen. (4) Gefunctionaliseerd edialkylzinkreagentia geven nagenoeg dezelfde hoge regio- en enantioselectiviteit en(5) waarschijnlijk is steeds hetzelfde chirale kopercomplex verantwoordelijk voor dewaargenomen enantioselectiviteiten en configuraties van de 1,4-produkten. Aangezien de combinatie van het fosforamidaat en koper heeft geleid tot een snelle eneffectieve katalyse van de 1,4-additie van dialkylzinkreagentia aan ",$-onverzadigdeketonen, is deze combinatie getest als chirale katalysator in allerlei andere koolstof-koolstof binding vormende reakties. De resultaten van deze inleidende experimentenstaan kort beschreven in hoofdstuk 7.

university of groningen catalytic enantioselective conjugate … · 2016-03-08 · damage.4 the...

Documents