University of Groningen

Asymmetric catalysis in the synthesis of cis-cyclopropyl containing fatty acids and the additionof Grignard reagents to carbonyl compoundsHanstein, Miriam

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:2014

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Hanstein, M. (2014). Asymmetric catalysis in the synthesis of cis-cyclopropyl containing fatty acids and theaddition of Grignard reagents to carbonyl compounds. [S.l.]: [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: 19-05-2020

Asymmetric catalysis in the synthesis of cis-cyclopropyl containing fatty acids and

the addition of Grignard reagents to carbonyl compounds

Miriam Hanstein

The work described in this thesis was carried out at the Stratingh Institute for Chemistry, University of Groningen, The Netherlands. This work was financially supported by the University of Groningen and the Netherlands Organisation for Scientific Research. Printed by Ipskamp Drukkers BV, Enschede, The Netherlands Cover design by Miriam Hanstein Cover art by Christo (big air package) ISBN: 978-90-367-6761-3 (printed version) ISBN: 978-90-367-6762-0 (digital version)

Asymmetric catalysis in the synthesis of cis-cyclopropyl containing fatty acids and the addition of Grignard reagents

to carbonyl compounds Proefschrift

ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen

op gezag van de rector magnificus, prof. dr. E. Sterken,

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

vrijdag 14 februari 2014 om 12.45 uur door

Miriam Hanstein

geboren op 14 april 1980 te Unna, Duitsland

Promotor: Prof. dr. ir. A.J. Minnaard Beoordelingscommissie: Prof. dr. S.R. Harutyunyan Prof. dr. ir. H.J. Heeres Prof. dr. R.J. Pieters

To Ulrike & Wolfgang

Table of Contents

Chapter 1 Introduction

1.1 Cyclopropyl containing fatty acids 2

1.2 Copper(I)-catalyzed 1,2-addition of Grignard reagents to ketones 9

1.3 Outline of this thesis 15

1.4 References 16

Chapter 2 Asymmetric synthesis of a cis-configured cyclopropyl

building block

2.1 Introduction 20

2.2 Reported syntheses of mycolic acids 21

2.3 Retrosynthetic analysis 25

2.4 Results and discussion 28

2.5 Conclusion 31

2.6 Experimental 32

2.7 References 37

Chapter 3 Mimicking the natural cell membrane for

mechanosensitive channels

3.1 Introduction 40

3.2 Reported syntheses of lactobacillic acid 41

3.3 Retrosynthetic analysis 45

3.4 Results and discussion 47

3.5 Conclusion 55

3.6 Experimental 55

3.7 References 64

Chapter 4 Copper(I)-catalyzed asymmetric alkylation of

aldehydes with Grignard reagents

4.1 Introduction 68

4.2 Goal of this study 71

4.3 Results and discussion 72

4.4 Conclusion 79

4.5 Experimental 80

4.6 References 85

Chapter 5 Enantioselective copper(I)-catalyzed alkylation of aryl

alkyl ketones with Grignard reagents

5.1 Introduction 88

5.2 Goal of this study 90

5.3 Results and discussion 91

5.4 Conclusion 95

5.5 Experimental 96

5.6 References 104

Summary 105

Samenvatting 111

Zusammenfassung 117

Acknowledgements 123

Introduction

This chapter gives an overview of cyclopropyl containing fatty acids found in nature. In addition, their biosynthetic pathways and possible degradation processes are discussed. In the second part, the state of the art in copper(I)-catalyzed 1,2-addition of Grignard reagents to ketones is summarized.

Chapter 1

2

1.1 Cyclopropyl containing fatty acids

In nature cyclopropyl rings are found in many compounds. In fatty acids, both cis- and trans-configurations do occur. An example of a trans-cyclopropyl containing fatty acid, lyngbyoic acid, was lately found in a marine cyanobacterium, in addition to other examples isolated from different marine cyanobacteria.1–3 This part of the introduction, however, will focus on fatty acids containing the more common cis-cyclopropane moiety, found in both prokaryotes and eukaryotes. Some of the major compounds commonly isolated are depicted in Figure 1.1.4 One of the first examples of a cis-cyclopropane containing fatty acid, lactobacillic acid, was found in Lactobacillus arabinosus by Hofmann et al. already in 1950.5 Later, this fatty acid and its homologs were also isolated from plants.6,7 Cis-cyclopropane fatty acids are usually esterified as an acyl chain of phospholipids, and mainly located in the cell membrane.

In plants, cis-cyclopropane containing fatty acids are mostly isolated from seed oils, but also found in vegetative plant parts (e.g. leaves, and roots).8 Dehydrosterculic acid was characterized as a major fatty acid in most plant families, e.g.: Bombacaceae, Malvaceae, Steruliaceae, Leguminosae and Rannunculaceae.6 This acid has been proposed as the precursor in the biosynthetic pathway of sterculic acid, a cis-cyclopropene fatty acid. Different investigations showed, that cis-cyclopropane fatty acids have a biological and physiological effect on animals.9 The biological effects can be interpreted as a defense mechanism to fungal attack. Sterculic acid was noticed as a potent inhibitor of mammalian 9 desaturases.10 In addition cis-cyclopropane fatty acids may function as carbon and energy storage in seeds.11

The production of cis-cyclopropane fatty acids has been observed in both gram-positive and gram-negative bacteria.9 Here, the predominantly isolated cis-cyclopropane containing fatty acids are lactobacillic acid, dehydrosterculic acid and cis-9,10-methylenehexadecanoic acid (Figure 1).4 A special type of cis-cyclopropane fatty acids, mycolic acids, were detected in the cell membrane of mycobacteria.12 These fatty acids have two cyclopropyl moieties and depending on the type, additional functional groups, e.g. methoxy- or keto- groups, next to the cyclopropyl ring (for more details see chapter 2). During bacterial growth, cis-cyclopropane fatty acids were shown to be formed mainly in the late exponential/early stationary phase.13 The interest to understand the transformation from unsaturated fatty acids to cis-cyclopropane fatty acids, increased with the relevance to approach this biosynthesis as a means to inhibit bacterial growth.14

Introduction

3

OH

O

OH

O

OH

O

OH

O

OH

O

OH

O

OH

O

OH

O

OH

O

OH

O

cis-3,4-methylenedodecanoic acid

cis-4,5-methylenetridecanoic acid

cis-5,6-methylenetetradecanoic acid

cis-6,7-methylenepentadecanoic acid

cis-7,8-methylenehexadecanoic acid

cis-8,9-methyleneheptadecanoic aciddihydromalvalic acid

cis-9,10-methyleneoctadecanoic aciddihydrosterculic acid

cis-11,12-methyleneoctadecanoic acidlactobacillic acid

cis-3,4-methylenedecanoic acid

cis-9,10-methylenehexadecanoic acid

cis-9,10-methyleneheptadecanoic acid

13:CA n-9

14:CA n-9

15:CA n-9

16:CA n-9

17:CA n-9

18:CA n-9

19:CA n-9

11:CA n-7

17:CA n-7

18:CA n-8

19:CA n-7

OH

O

number of carbons: cyclopropyl moiety (CA) position of ring (n-x)

OH

COOH

mycolic acid

Figure 1.1. Overview of cis-cyclopropane fatty acids found in bacteria and plants.

Chapter 1

4

The appearance of cis-cyclopropane fatty acids was detected, when bacteria needed to resist environmental stress in the form of an acidic shock. One model proposes the decrease of proton permeability by transformation of unsaturated fatty acids to cis-cyclopropane fatty acids in the lipid component. This phenomenon was investigated for the microorganism Helicobacter pyleri, which is causing gastritis and peptic ulcer disease.13

The influence on the physical properties, e.g. the transition temperature, after conversion of an unsaturated fatty acid to a cis-cyclopropane fatty acid was shown to be modest.15 The drug resistance of Mycobacterium tuberculosis is thought to be due to the low permeability affected by mycolic acids as main component of the cell envelope.16 This suggests, that cis-cyclopropane fatty acids play a role in the physiology of pathogenesis. For example it was observed, that during the stationary phase, when more cis-cyclopropane fatty acids are accumulated in the cell wall, E. coli shows a higher resistance against hyperbaric oxygen treatment.17,18 Still the physiological analysis remains difficult, therefore the enzymes involved in the biotransformation are intensively studied.

The transformation of unsaturated phospholipids to the corresponding cyclopropyl containing lipids was proposed to be catalyzed by cyclopropyl fatty acid (CFA) synthase, and it is postulated that the reaction takes place in or close to the cell membrane.4,13 It is generally accepted that the methylene group transferred to the C-C double bond derives from S-adenosyl methionine 1 (Scheme 1.1). This hypothesis was investigated by using an analog of 1, where two protons of the methyl group are substituted by deuterium and tritium (Scheme 1.1).19,20 In the first step, the -electrons of the double bond act as nucleophile and attack sulfonium cation 1 in an SN2 fashion. This gives carbocation 2 as an intermediate, and an inversion of the stereochemistry of the methyl group is observed. After cyclization and proton abstraction, the cyclopropane ring is formed. Subsequent degradation experiments and analysis of the stereoisomers confirmed the stereochemistry of the reaction. The cis-configuration of the cyclopropane fatty acid, formed in vivo, showed to be higher than 99%.13 The absolute configuration of the cyclopropyl moiety of lactobacillic acid, derived e.g. from E. coli or Lactobacillus plantarum, was identified as R/S by comparison of optical rotations of keto derivatives.20,21

Introduction

5

1112

SN2H D

TB TD

1112

BH

H+

TD

12

O

TD

11

O

Me S CO2

Ad

NH3

Ad: adenosylS-adenosyl methionine 1

2

RS

Ad

H TD

Scheme 1.1. Mechanism and evaluation of the stereochemistry of the methyl transfer from S-adenosyl methionine to the C-C double bond in fatty acids.

Similar labeling studies in various Malva species supported the hypothesis, that sterculic acid derives from oleic acid with dihydrosterculic acid as an intermediate.22 The biosynthetic pathway showed above was established for both bacteria and plants.

Until now, the gene which encodes for CFA synthase, has been identified in E. coli, M. tuberculosis (bacteria) and S. foetida (eukaryote, plant).11 The comparison between the different amino acid sequences of the gene showed a highly conserved region at the C terminus. A significant difference was found in the length of the amino acid sequence, the one identified in bacteria is only half the length of the one of S. foetida.

Experiments with the bacterial CFA synthase in E. coli, demonstrated a substrate specificity for cis-9, cis-10 and cis-11 unsaturated fatty acids. However, cis-double bonds close to the end of the C18 fatty acyl chain were not transformed into the corresponding cis-cyclopropane fatty acid.13 In addition no cis-cyclopropane fatty acid formation was observed, when the double bond of the fatty acid was trans configured. In general cis-cyclopropane fatty acids were shown to be mainly esterified at the sn-2 position of phospholipids bearing a phosphatidylethanolamine head group. Such a specification, has not yet been identified for the CFA synthase form S. foetida.11

For both plants and bacteria the highest CFA synthase activity was observed close to the cell membrane, indicating that the formation of the cyclopropyl moiety is a postsynthetic modification of phospholipids.15 Since no transmembrane domain was found in the amino acid sequence of S. foetida, it was proposed that the enzyme is either a membrane-associated or an integral membrane protein.11 In E. coli the CFA

Chapter 1

6

synthase has been isolated from the cytoplasmic fraction, and showed to be only stabilized in vesicles containing unsaturated fatty acids or cyclopropane fatty acids. In addition it was observed, that the CFA synthase is able to cyclopropanate inner and outer leaflets of unilamellar vesicles. This implies a mechanism, in which the enzyme flips an acyl chain out of the lipid bi-layer, performs the cyclopropanation, and releases the chain back to the lipid bilayer.13,23

Alternative pathways, after the formation of intermediate 2, have been postulated as depicted in Scheme 1.2.15 Following a 1,2-hydride shift to the more stable carbocation 3 and subsequent 1,2-elimination, a methylene group 4 is obtained. A direct 1,2-elimination gives an isoprene unit 5. The cyclopropyl moiety 6 is formed after a 1,3-proton elimination reaction. An allylic methyl branched unit 7 is observed after 1,2-elimination. Finally, the capture of water by intermediate 2 results in compound 8, which generates a hydroxyl group after proton expulsion.

RS

Ad

CH3

SN2

H3C1,2 hydride shiftH3C

H+ H+H+

H3C CH3

H+

+H2O H3C OH2

H+

CH3

OH

23

4 5 6 7 9

8

Scheme 1.2. Possible products derived from intermediate 2.

The degradation and metabolism of cyclopropane moieties in natural products has been studied much less than the biosynthetic pathway. Most commonly, ring-opening reactions have been observed. These reactions are thermodynamically favoured and de facto irreversible.24

Introduction

7

In in vivo experiments synthetic cis/trans isomer mixtures of 9,10-methylene octadecanoate (9,10-cyclopropyl octadecanoate) were fed to rats, to study the metabolism of cyclopropane fatty acids. The analysis of lipid extracts from adipose tissue, showed accumulation of the 3,4-methylene dodecanoic acid of both isomers. This indicated the inability of further -oxidation down the fatty acid chain.25 A similar observation was made, when rat-liver mitochondria were incubated in vitro with radio-labeled [9,10-methylene-14C]-cis-9,10-methylene hexadecanoic acid. As degradation product radio-labeled 14CO2 was expected, but instead cis-3,4-methylene decanoic acid was isolated and identified by IR analysis.26

However, complete degradation of cyclopropane fatty acids has been noticed for example in Ochromonas danica (eukaryote, algea) and Tetrahymena pyriformis (unicellular eukaryote).27,28 In both studies, cis-cyclopropane fatty acids with a 14C isotope labeled at the methylene carbon of the cyclopropyl ring were incubated with the respective organism. In both cases 14CO2 was detected, indicating complete degradation of the cis-cyclopropane fatty acid, since the isotope was selectively introduced in the cyclopropyl ring. Based on these results Tipton and co-workers postulated the following metabolism pathway of oxidation (Scheme 1.3).

Chapter 1

8

CoA

O*

CoA

* O

CoA

O

+ 4 CH3CO CoA

CoA

O

OH

*

*

* CoA

OO

CoA*

O+ CH3CO CoA

CH3CH2CO CoA + 2 CH3CO CoA + *CH3CO CoA

Scheme 1.3. Proposed mechanism for the degradation of cis-cyclopropane fatty acids.

An enzyme which catalyzes several oxidative transformations, is cytochrome P450, a member of the superfamily of heme-dependent mono-oxygenases. In the proposed mechanism of P450 mediated hydroxylation, hydrogen abstraction leads to radical intermediates, e.g. a radical to the cyclopropyl ring in cis-cyclopropane fatty acids. The oxidative rearrangement was studied with cis-cyclopropane fatty acid 10 as substrate (Scheme 1.4).29 Depending on the ring scission, primary alcohol 12 or secondary alcohol 13 could be obtained. However, the experimental data showed

Introduction

9

formation of alcohol 12 and -hydroxylated product 14. Those results demonstrated the formation of a radical intermediate in the ring-opening reaction.

OH

O

[Fe=O]3+

OH

O

[Fe-OH]3+

OH

O

OH

10

11

14

[Fe-OH]3+

A

B

[Fe-OH]3+

A

B

OH

O

12

CH2OH

OH

O

13

OH

Scheme 1.4. Possible hydroxylation products of cis-cyclopropane fatty acid 10.

In summary, cis-cyclopropyl fatty acids have been studied from different points of view, but still many cell processes are not completely understood. The use of synthetic, enantiopure cis-cyclopropyl fatty acids in in vivo studies, with additional labeling, will be useful to follow the cellular pathways e.g. in M. tuberculosis.

1.2 Copper(I)-catalyzed 1,2-addition of Grignard reagents to ketones

In this part of the introduction, an overview of a new asymmetric 1,2-addition methodology, developed in our institute, is given.30–35 Secondary and tertiary alcohols are general motifs found in many natural products, pharmaceuticals and crop protection products. One of the most straightforward approaches to synthesize compounds with these functional groups, is the addition of organometallic reagents to aldehydes or ketones.36 Although for aldehydes the field is dominated by the catalyzed addition of dialkylzinc reagents, also Grignard reagents in the presence of chiral titanium-complexes have been used, and this combination of reagents leads to high yields and enantioselectivities.37–39 For the synthesis of chiral tertiary alcohols, considerably less research has been performed, still the described examples use titanium-complexes as well.40,41 In both cases, the major drawback of this method is the required stoichiometric amount of titanium. Therefore, the development of a new

Chapter 1

10

catalytic system, in which transmetallation to titanium would be avoided, was a challenge. The idea to develop a methodology based on a chiral copper(I)-complex arose, when Lipshutz and co-workers reported a highly regioselective 1,2-reduction with Cu-H for -unsaturated ketones.42

-unsaturated ketone 15b was used as standard substrate for the screening (Scheme 1.5). Ligand L1 gave the highest yield and enantioselectivity among the tested ferrocene, biaryl-based and phosphoramidite ligands. Different copper sources were investigated, and the best result was obtained using 2 as metal precursor. In terms of regio- and enantioselectivity, methyl tert-butyl ether (MTBE) was the solvent of choice.30

R2

O

R1

5 mol% CuBr.SMe26 mol% L11.3 eq iBuMgBrMTBE, 78 °C, 12 h R2

HO

R1

iBu

R2

O

R1

iBu+ +

R2

OH

R1

1,2-product 1,4-product 1,2-reduction

FePP

H3C H

L1rev-JosiPhos

R1: H 15a Me 15b Br 15c

16 17 18

R2: Me

Scheme 1.5. Possible regioisomeres formed during the reaction.

-unsaturated ketones (15a, 15b, 15c) were studied varying the substituent at the -position (R1) and the Grignard reagent (Table 1.1). For both substrates 15b, 15c, high regioselectivities were observed. The reactions in which the branched iso-butylmagnesium bromide was added, yielded higher enantioselectivities compared to the reactions where ethylmagnesium bromide was used. Due to these findings, further investigations focused on -unsaturated ketones possessing an -substituent and branched Grignard reagents.31

Introduction

11

Table 1.1. The influence of the substituent in -position on the regioselectivity and the influence of the Grignard reagent on the enantioselectivity.

Entry Substrate Grignard reagent 16:17:18 [%] 16, ee [%]

1 15a EtMgBr 16:84:0 14

2 15a iBuMgBr 51:49:0 32

3 15b EtMgBr 97:2:1 40

4 15b iBuMgBr 98:1:1 84

5 15c EtMgBr 98:1:1 42

6 15c iBuMgBr 97:1:2 90

In general for all substrates studied, a high isolated yield in the range of 81-96% was obtained. The best results in terms of enantioselectivity were observed for substrates, in which R2 is a methyl group (Figure 1.2). Additionally, one can see that for all branched Grignard reagents a high enantioselectivity is reached. This makes the new methodology complementary to the existing ones, mainly the addition of linear dialkylzincs.

HO HO HO HO

yieldee

95%40%

93%84%

95%92%

95%88%

HO

Br

HO

Br

HO

Br

HO

Br

yieldee

96%39%

94%90%

94%96%

96%94%

Figure 1.2. The addition of various Grignard reagents to substrate 15b or 15c.

The tertiary alcohols formed from substrate 15c, offer the possibility to synthesize two different types of dihydrofurans 19 and 20 and cyclopentenol 21 (Scheme 1.6). Following route A, implies that the -bromo substituent reacts with phenylacetylene in a Sonogashira reaction to the corresponding enynes. After treatment with a strong base, the cyclization product 3-benzylidene-2,3-dihydrofuran 19 is obtained. 2,5-Dihydrofurans 20 can be synthesized via route B, from alcohol 15c by debromination, subsequent alkylation of the hydroxyl-group with allyl bromide and

Chapter 1

12

ring-closing metathesis of the diene. To form cyclopentenol 21, route C was followed, here butenylmagnesium bromide was used as Grignard reagent in the 1,2-addition, followed by debromination as in route B and as final step ring-closing metathesis catalyzed by Grubbs second generation catalyst.43

R1

Br

R2HOR1

OH

C O R2

R1Ph

Ph

O R2

R1

A

B

15c19

21

20

Scheme 1.6. Novel approaches for the synthesis of dihydrofurans and cyclopentenols.

-unsaturated ketones as substrates, the methodology was extended and applied to aryl alkyl ketones. For this study, mainly substituted acetophenones were used (Scheme 1.7).32 By varying the position of the substituent on the phenyl ring, it was attempted to discover a relation between steric and electronic effects on the enantioselectivity. Surprisingly, no clear trend was noticed. The only consistency was that for the substrates with substituents in para or meta position a higher ee was obtained. In addition one can note a lower isolated yield for substrates with an ortho substituent or more bulk close to the carbonyl group, as in the case of 23c, 23f, and 23g.

Several other Grignard reagents were added to 3,5-ditrifluoromethyl acetophenone (substrate not shown), and again the branched Grignard reagents showed the highest enantioselectivity. The lowest ee, 22%, was observed when ethylmagnesium bromide was used.

Introduction

13

O

EtMgBr

Et+

5 mol% CuBr.SMe26 mol% L1MTBE, 78 °C, 12 h HO

Et

Et

23R R

22

HOEt

EtHO

Et

Et

BrBr

HOEt

Et

Br

HOEt

EtHO

Et

Et

FF

HOEt

Et

CF3

F3C

HOEt

EtHOEt

Et

yieldee

91%86%

96%90%

61%95%

95%98%

yieldee

95%76%

84%70%

71%95%

91%66%

23a 23b 23c 23d

23e 23f 23g 23h

Scheme 1.7. Asymmetric 1,2-addition of 2-ethylbutylmagnesium bromide to aryl ketones.

The transition state proposed for the 1,2-addition differs for aryl alkyl ketones and -unsaturated ketones, but for both cases first the alkyl group is transferred to the

copper to form a chiral alkyl-Cu(I)-complex (Scheme 1.8).33 The proposal is that this copper-species forms with aryl alkyl ketones a pseudo-chair transition state, where the copper coordinates to the double bond of the carbonyl group. In the transition state with -unsaturated ketones as substrates, the copper species forms a

-complex with the carbon-carbon double bond. In both cases, the magnesium salt functions as a Lewis acid and activates the carbonyl group.

Chapter 1

14

CuP

P Br

BrCu

P

P+ RMgBr

CuP

P R

BrMgBr

Ar

O

Me

O

R2

XR1

O

R2

XR1

CuPP

MgBrBr

OMe

PhMg

Br

Br

CuP P

R

R2

XR1

R OMgBr

Ar MeR OMgBr

R

Scheme 1.8. Proposed transition states for the copper-catalyzed 1,2-addition.

For a number of catalytic enantioselective alkylation reactions with different organometallic reagents to either aldehydes or ketones, a non-linear effect has been observed.44 This means that the ee of the product is not correlated in a linear fashion with the ee of the catalyst. In the case, in which the ee is higher than the value of the corresponding linear value, this effect is termed a positive non-linear effect. As a result of this, the enantioselectivity of the product is amplified compared to the ee of the catalyst. This phenomenon was noticed in the 1,2-addition of (2-ethylbutyl)magnesiumbromide to substrate 15c (Table 1.2).34 The scalemic copper-complex was formed in situ in two ways: 1) both enantiomers of the rev-Josiphos ligand L1 were mixed together with the corresponding amount of copper salt and stirred for 30 min; 2) solutions of each enantiomer as Cu-complex were prepared and those were mixed to the corresponding solutions and stirred for 30 min. While preparing the racemic solutions of the complex, independent of the preparation method, within 20-30 min significant amounts of precipitate were observed. Therefore, all the reactions in Table 1.2 were performed twice: one time with the mixture of supernatant and precipitate (sp) and the other time only with the supernatant (s). The results for both ways are very similar, the ee is around 5% higher, when only the supernatant was used.

Introduction

15

The asymmetric amplification observed in this study, can be explained by the different solubility of the homochiral dimer and the heterochiral dimer of the copper-complex in MTBE. The homochiral dimer stays in the mother liquor, while the heterochiral dimer precipitates.34

Table 1.2. Asymmetric amplification in the 1,2-addition of Grignard reagent to enone 15c.

EtMgBr

Et+

Me

O

BrMe

HO

Br

5 mol% L1/CuBr.SMe2MTBE, 78 °C, 24 h

15c 16c

entry ee catalyst loading 16c eesp conversion 16c ees conversion 1 20 % 25 % 80 % 75 % 94 % 92 % 2 40 % 12 % 90 % 82 % 94 % 95 % 3 60 % 8 % 90 % 93 % 92 % 92 % 4 80 % 6 % 90 % 90 % 94 % 93 %

eesp: reactions catalyzed by a supernatant-precipitate mixture; ees reactions catalyzed by supernatant.

1.3 Outline of this thesis

This thesis is divided in two parts, in the first two chapters, different synthetic approaches towards cyclopropyl containing fatty acids are studied. In the second part of the thesis, the application of the lately in our institute developed copper(I)-catalyzed addition of Grignard reagents to aryl aldehydes and aryl ketones is described.

Chapter 2 describes the efforts to synthesize an enantiopure -mycolic acid, which has been isolated from M. tuberculosis. It was tried to synthesize a common building block for both cyclopropyl moieties in the meromycolate chain, using rhodium-catalyzed cyclopropanation.

In chapter 3, the synthesis of lactobacillic acid, a cyclopropyl fatty acid mostly found in the cell membrane of bacteria, is discussed. For the synthesis of the cis-configured cyclopropyl ring, we envisioned a reaction sequence, in which the hetero asymmetric allylic alkylation and RCM are the key steps to build a chiral internal allylic alcohol.

Chapter 1

16

In chapter 4, the results of the application of the novel copper(I)-catalyzed 1,2- -unsaturated aldehydes are presented. The reactions performed using the standard conditions, show significantly lower enantioselectivities than the corresponding aryl ketones.

The topic of chapter 5 is the broadening of the substrate scope in the copper-catalyzed alkylation of aryl alkyl ketones with Grignard reagents. In addition, the relation between enantioselectivity and steric and electronic effects is studied.

1.4 References

1. N. Sitachitta and W. H. Gerwick, J. Nat. Prod., 1998, 61, 681–684.

2. J. B. MacMillan and T. F. Molinski, J. Nat. Prod., 2005, 68, 604–606.

3. J. C. Kwan, T. Meickle, D. Ladwa, M. Teplitski, V. Paul, and H. Luesch, Mol. Biosyst., 2011, 7, 1205–1216.

4. W. W. Christie, Topics in lipid chemistry, Vol. I, 1970, 1-49.

5. K. Hofmann and R. A. Lucas, J. Am. Chem. Soc., 1950, 72, 4328–4329.

6. L. O. Hanus, P. Goldshlag, and V. M. Dembitsky, Biomed. Pap., 2008, 152, 41–45.

7. E. M. Gaydou, A. Ralaimanarivo, and J. P. Bianchini, J. Agric. Food Chem., 1993, 41, 886–890.

8. I. Yano, B. W. Nichols, L. J. Morris, and A. T. James, Lipids, 1972, 7, 30–34.

9. G. A. Nolen, J. C. Alexander, and N. R. Artman, J. Nutr., 1967, 93, 337–348.

10. A. C. Fogerty, A. R. Johnson, and J. A. Pearson, Lipids, 1972, 7, 335–338.

11. X. Bao, S. Katz, M. Pollard, and J. Ohlrogge, Proc. Natl. Acad. Sci., 2002, 99, 7172–7177.

12. C. E. Barry III, R. E. Lee, K. Mdluli, A. E. Sampson, B. G. Schroeder, R. A. Slayden, and Y. Yuan, Prog. Lipid Res., 1998, 37, 143–179.

13. D. W. Grogan and J. E. Cronan, Microbiol. Mol. Biol. Rev., 1997, 61, 429–441.

14. Y. Yuan, R. E. Lee, G. S. Besra, J. T. Belisle, and C. E. Barry, Proc. Natl. Acad. Sci., 1995, 92, 6630–6634.

15. P. H. Buist, Nat. Prod. Rep., 2007, 24, 1110–1127.

Introduction

17

16. J. P. Cegielski, Clin. Infect. Dis., 2010, 50, S195–S200.

17. J. B. Harley, G. M. Santangelo, H. Rasmussen, and H. Goldfine, J. Bacteriol., 1978, 134, 808–820.

18. K. M. George, Y. Yuan, D. R. Sherman, and C. E. Barry, J. Biol. Chem., 1995, 270, 27292–27298.

19. S. Rasyoni, Diss. ETH, # 11318, 1995.

20. L. J. Stuart, J. P. Buck, A. E. Tremblay, and P. H. Buist, Org. Lett., 2006, 8, 79–81.

21. G. D. Coxon, J. R. Al-Dulayymi, M. S. Baird, S. Knobl, E. Roberts, and D. E. Minnikin, Tetrahedron Asymmetry, 2003, 14, 1211–1222.

22. I. Yano, L. J. Morris, B. W. Nichols, and A. T. Jams, Lipids, 1972, 7, 35–45.

23. F. R. Taylor and J. E. Cronan, Biochemistry (Mosc.), 1979, 18, 3292–3300.

24. L. A. Wessjohann, W. Brandt, and T. Thiemann, Chem. Rev., 2003, 103, 1625–1648.

25. R. Wood and R. Reiser, J. Am. Oil Chem. Soc., 1965, 42, 315–320.

26. A. E. Chung, Biochim. Biophys. Acta, 1966, 116, 205–213.

27. C. L. Tipton and N. M. Al-Shathir, J. Biol. Chem., 1974, 249, 886–889.

28. W. J. Magat and C. L. Tipton, Arch. Biochem. Biophys., 1970, 141, 353–355.

29. M. J. Cryle, P. R. Ortiz de Montellano, and J. J. De Voss, J. Org. Chem., 2005, 70, 2455–2469.

30. A. V. R. Madduri, A. J. Minnaard, and S. R. Harutyunyan, Chem. Commun., 2012, 48, 1478–1480.

31. A. V. R. Madduri, A. J. Minnaard, and S. R. Harutyunyan, Org. Biomol. Chem., 2012, 10, 2878–2884.

32. A. V. R. Madduri, S. R. Harutyunyan, and A. J. Minnaard, Angew. Chem. Int. Ed., 2012, 51, 3164–3167.

33. A. V. R. Madduri, S. R. Harutyunyan, and A. J. Minnaard, Drug Discov. Today Technol, 2013, 10, e21-e27.

34. F. Caprioli, A. V. R. Madduri, A. J. Minnaard, and S. R. Harutyunyan, Chem. Commun., 2013, 49, 5450–5452.

Chapter 1

18

35. Z. Wu, A. V. R. Madduri, S. R. Harutyunyan, and A. J. Minnaard, Eur. J. Org. Chem., 2013, DOI: 10.1002/ejoc.201301476.

36. R. Noyori and S. Hashiguchi, in Applied Homogeneous Catalysis with Organometallic Compounds, eds. B. Cornils and W. A. Herrmann, Wiley-VCH Verlag GmbH, 2008, 552–571.

37. D. Seebach, L. Behrendt, and D. Felix, Angew. Chem. Int. Ed., 1991, 30, 1008–1009.

38. E. Fernández-Mateos, B. Maciá, and M. Yus, Adv. Synth. Catal., 2013, 1249-1254.

39. E. Fernández-Mateos, B. Maciá, D. J. Ramón, and M. Yus, Eur. J. Org. Chem., 2011, 2011, 6851–6855.

40. S.-J. Jeon, H. Li, C. García, L. K. LaRochelle, and P. J. Walsh, J. Org. Chem., 2005, 70, 448–455.

41. M. Yus, D. J. Ramón, and O. Prieto, Tetrahedron Asymmetry, 2002, 13, 2291–2293.

42. J. Am. Chem. Soc., 2010, 132, 7852–7853.

43. Y. Muramatsu and T. Harada, Angew. Chem. Int. Ed., 2008, 47, 1088–1090.

44. T. Satyanarayana, S. Abraham, and H. B. Kagan, Angew. Chem. Int. Ed., 2009, 48, 456–494.

Asymmetric synthesis of a cis-configured cyclopropyl building block: studies towards the synthesis of mycolic acid

In this chapter our efforts to synthesize an -mycolic acid, a main component found in the cell wall of Mycobacterium tuberculosis, are described. For the cyclopropyl moieties in the mycolic acid, the synthesis of a common building block via intramolecular cyclopropanation of allyl diazoactetate was studied. For the -hydroxy ester motif we aimed to apply an asymmetric hydrogenation to introduce enantioselectivity.

Chapter 2

20

2.1 Introduction

In 2009, the WHO (World Health Organization) published an up-date of the global control of tuberculosis (TB): each year over nine million new cases are reported, with nearly two million deaths annually caused by infection by Mycobacterium tuberculosis (M. tuberculosis).1 In the last decade, M. tuberculosis strains that are multidrug-resistant and even extensively-drug resistant have been found, therefore the need to understand this resistance becomes urgent.2

The persistence of M. tuberculosis can partly be explained by the low permeability of the cell wall for antibiotics and other drugs. The impermeability is found in most mycobacteria and caused by a thick and robust outer membrane.3,4 The highly complex cell envelope consists of three structural components: the plasma membrane, the cell wall core and the capsule (Figure 2.1).5,6

Figure 2.1. Schematic structure of the cell envelope of M. tuberculosis.

The main compounds forming the cell wall (40-60% of the cells dry weight) are mycolic acids 1 (Figure 2.2), which are either found as free fatty acids, as monoesters or as diesters of trehalose, glucose and glycerol or covalently bound to the arabinogalactan layer.7,8 The initial evaluation and characterization of mycolic acids from M. tuberculosis was reported by Lederer et al. and later Minnikin and co-workers described the detailed structure of the major classes of mycolic acids, including their stereochemistry.6,9 Mycolic acids are -branched -hydroxyl fatty acids with up to ninety carbons. A mycolic acid can be divided into two parts; an un-functionalized mycolic chain and a meromycolate chain with up to two functional groups [X] and

Studies towards the synthesis of mycolic acid

21

[Y] (Figure 2.2). Depending on the functional group in the distal position [X] one can distinguish between the three major classes of mycolic acids: -mycolic acids, methoxy-mycolic acids and keto-mycolic acids.

Me

OMe

Me

O

Me

[X] [Y] OH

OH O

a c db

mycolic chainmeromycolate chain

[X]

- methoxy- keto-

[Y] or orn

orMe

or

a: 15, 17, 18, 19 b: 10,14, 16c: 11, 15, 17, 19, 21 d: 21,23

mycolic acid 1

Figure 2.2. Major types of mycolic acids from M. tuberculosis.

Watanabe et al. reported the detailed composition of mycolic acid components of different M. tuberculosis strains and other mycobacteria species. Analysis of the cell wall extract by TLC gives rise to a specific pattern for each Mycobacterium species.10 This characteristic makes mycolic acids attractive as straightforward diagnostic markers for mycobacterial infections. Different groups published the possibility to detect mycolic acids by either HPLC or mass spectra analysis in sputum samples of possible TB infected patients.11,12 Thus, we envisioned that synthesized, (enantiopure) mycolic acids could be employed as an internal standard in those measurements. With such an internal standard the amount of M. tuberculosis in the sputum can be quantified, as well the identification of the right fractions isolated from the sputum would be easier.

2.2 Reported syntheses of mycolic acids

In the field of mycolic acid synthesis, two research groups have been mainly involved: Baird et al. and Minnikin et al.. These groups were able to synthesize examples of all the different mycolic acid classes,13–16 but since -mycolic acids are

Chapter 2

22

the major component of the cell wall of M. tuberculosis, we focused only on the synthesis of this mycolic acid.17

The first step of their reported synthesis is the desymmetrization reaction of meso diester 2 to monoester 3 with pig liver esterase at pH 6.5 (Scheme 2.1). The free hydroxyl group is oxidized with PCC to aldehyde 4, which was connected to an alkyl chain in a Wittig reaction giving olefin 5 with 81% yield and a Z/E ratio of 6:1.

Me(CH2)18PPh3Br BuLi, THF

pig liver esterase ethylene glycol, water, pH 6.5

2

O Pr

O

OPr

O3

O Pr

O

HO80%

4

O Pr

O

O

PCC, CH2Cl288%

H81%

6:1 (Z/E)5

O Pr

O

H3C(H2C)17

1) LiAlH4, THF2) N2H4, NaIO4 AcOH, CuSO4, i-PrOH3) PCC, CH2Cl2

68%

H3C(H2C)19

6

O

H(CH2)12CHOH3C(H2C)19

7

8

O Pr

O

SO

ON

S

LiHMDS

(CH2)12H3C(H2C)19

9

CH2OCOPr

43%

77%

1) LiAlH4, THF2) N2H4, NaIO4 AcOH, CuSO4, i-PrOH

(CH2)14H3C(H2C)19

10

OH

Scheme 2.1. Synthesis of the meromycolate chain in an -mycolic acid.

In the subsequent steps, the ester is reduced to the corresponding alcohol, the double bond is reduced under mild conditions because of the presence of the cyclopropyl unit, and the free alcohol is oxidized to give aldehyde 6. Then again the alkyl chain is

Studies towards the synthesis of mycolic acid

23

extended via Wittig reaction with an ester group at the end, applying the same repetitive steps as described before to get aldehyde 7. In a Julia-Kocienski olefination of aldehyde 7 with sulfone 8, the second cyclopropyl moiety of the meromycolate chain was installed. Sulfone building block 8 was made from the same starting material 2. Alcohol 10 was obtained using the same steps as before: reduction of the ester group to the corresponding alcohol and reduction of the double bond with diimide.

The -hydroxy ester part was prepared from epoxide 11 by ring opening with a Grignard reagent to give secondary alcohol 12 (Scheme 2.2). In the next steps the free alcohol was protected and the benzyl group was removed. After oxidation, the resulting acid was transformed into methyl ester 13, and under these acidic conditions the THP-group was removed.

O

BnOH

BrMg(CH2)9OTHPCuI, 2 h, 30 °C

BnO (CH2)10OTHP

OH

1) imidazole, DMF, tBuSiMe2Cl

2) H2, Pd/C, MeOH

3) NaIO4, RuCl3.H2O

CH3CN, H2O, CCl44) MeOH, H2SO4

MeO (CH2)10OH

OHO

11 12

13

1) tBuPh2SiCl, DMAP, Et3N2) LDA, CH3(CH2)23I, HMPA

86%

54%

26%MeO (CH2)10OSiPh2

tBu

OHO

14

(CH2)23CH3

1) Ac2O, pyridine2) F3) PCC

61%

MeO (CH2)9CHO

OAcO

15

(CH2)23CH3

Scheme 2.2. Synthesis of the -hydroxy ester moiety in mycolic acid, using epoxide 11.

Next, the free primary alcohol was protected with a TBDPS-group and in a Fráter alkylation the long mycolic chain was introduced. Then the secondary alcohol was

Chapter 2

24

protected, followed by deprotection of the primary alcohol and oxidation of this to give aldehyde 15.

In the final part of the synthesis, the meromycolate chain is attached to the -hydroxy ester part (Scheme 2.3), and in order to do this, alcohol 10 was converted to sulfone 16. In a Julia-Kocienski reaction the two parts were coupled to give olefin 17 as an E/Z mixture. Reduction of the double bond with diimide in situ generated from dipotassium azodicarboxylate led to the protected -mycolic acid 18.

MeO (CH2)9CHO

OAcO

15(CH2)23CH3

(CH2)14H3C(H2C)19

10

OH

S

NSH PPh3, DEAD1)

2) MCPBA, CH2Cl2

(CH2)14H3C(H2C)19

16

SO

OS

N

LiHMDS

(CH2)14H3C(H2C)19

17

(CH2)9OMe

OAc

(CH2)23CH3

O

Potassium azodicarboxylate,CH3COOH, THF

(CH2)14H3C(H2C)19

18

(CH2)11 OMe

OAc

(CH2)23CH3

O

Scheme 2.3. Coupling of the meromycolate chain to the -hydroxy ester part.

Studies towards the synthesis of mycolic acid

25

2.3 Retrosynthetic analysis

A closer look at the mycolic acid methyl ester 18 shows, that there are not many functional groups in the molecule. Therefore, it is challenging to connect the different building blocks in a limited number of steps (Scheme 2.4). For the coupling of the different fragments we preferably chose olefination reactions followed by diimide reduction, as this is tolerate by cyclopropyl moieties and also have been applied in the syntheses reported by Baird et al..17

In the first retro synthetic step, the mycolic chain is cleaved giving -hydroxy ester 19 (Scheme 2.4). We envisioned to obtain -hydroxy ester 19 via asymmetric hydrogenation of -keto ester 20. Subsequently the meromycolate chain is disconnected leading to alkyl bromide 21 and and -keto ester 22. The alkyl bromide can be divided into the aldehyde 23 and the sulfone building block 24.

Chapter 2

26

Scheme 2.4. Retrosynthetic analysis of the -hydroxy ester moiety with the meromycolate chain.

-my coli c acid est e r 18

1920

Studies towards the synthesis of mycolic acid

27

The meromycolate chain derivative 23 can be built from four building blocks: two alkyl chains and two cyclopropyl units (Scheme 2.5). Both cyclopropyl moieties can be prepared from the common cyclopropyl lactone 29, which is obtained in an intramolecular cyclopropanation reaction of allyl diazoacetate 30. For the distal cyclopropyl moiety, lactone 29 can be reduced and reaction of the resulting lactol with Wittig reagent 25 would lead to the first part of the chain. The free hydroxyl group can be transformed into a good leaving group and in a copper-catalyzed Grignard cross-coupling alkyl chain 26 can be linked. The alcohol terminus is subsequently deprotected and oxidized to the corresponding aldehyde. After protection of the hydroxyl group of building block 28 the ester can be reduced to an alcohol and converted into a sulfone group. Then the cyclopropyl building block can be connected in a Julia-Kocienski olefination to give meromycolate derivative 23 after selective oxidation of the free hydroxyl group.

23

O

O

O

OPPPh3Br

O OH

Br

OHO

O

25 26

27 28

29

O

O

30

H

N2

Scheme 2.5. Retrosynthetic analysis of the meromycolate chain.

We also considered the cobalt-catalyzed cyclopropanation described by Katsuki and co-workers that gives a building block similar to cyclopropyl unit 29. However, this approach was rejected due to difficulties in the synthesis of the required salen ligand, which is used as ligand in the cobalt-catalyzed cyclopropanation.18

Chapter 2

28

2.4 Results and discussion

2.4.1 Synthesis of the meromycolate chain

With this approach we tried to shorten the number of steps and searched for efficient reactions in comparison to the synthesis reported by Baird.17 In the process of designing a synthesis of enantiopure -mycolic acids, we required an asymmetric approach to prepare a cis-cyclopropyl building block. One option is an intramolecular cyclopropanation of allylic diazoacetates. We chose for the conditions described by Doyle et al., since the yield is high and an excellent enantioselectivity was reached (Scheme 2.6).19,20 In addition, the synthesis of Rh-catalyst 31 was shorter than that of the Co-salen catalyst Katsuki uses for the cyclopropanation.18 Lactone 29 obtained in this reaction, can be used as common building block for both cyclopropyl moieties of the meromycolate chain.

O

CHN2

O O O95%

0.1 mol% 31, CH2Cl2 reflux, 12-18 h

Rh2(4S-MEOX)4 31

30 29

Rh Rh

N O

O N COOMe

HMeOOC

H

OH

MeOOC

O N COOMe

H

Scheme 2.6. Intramolecular cyclopropanation as reported by Doyle et al..

Rh-catalyst 31 is synthesized in two steps starting from D or L-serine depending on the configuration of the cyclopropyl unit required. For our synthesis we started from L-serine (Scheme 2.7), and got oxazolidinone 33 in 69% yield and observed the same optical rotation as earlier reported. Next, ligand 33 was reacted with dirhodium(II) tetraacetate in refluxing chlorobenzene, which led to Rh2(4S-MEOX)4 by replacing the acetate ligands.21,22

HO NH3Cl

COOMe

69%O NH

O

COOMe

+ Rh2(OAc)4

NaCO3, sandchlorobenzene, reflux, 18 h

55%

(Cl3CO)2CO, Et3N, CH2Cl20 °C, 2 h, -78 °C, 30 min

32 3398% ee

34 31

Rh Rh

O N COOMeH

Scheme 2.7. Synthesis of Rh2(4S-MEOX)4 following Doyle’s procedure.

Studies towards the synthesis of mycolic acid

29

The starting material for the intramolecular cyclopropanation, diazoacetate 30, was formed in three steps from diketene acetone adduct 35 and allyl alcohol 36 (Scheme 2.8). After considerable optimization, -keto ester 37 was obtained in 72% yield. The two most significant changes to the literature procedure were to leave out the solvent and carry out the reaction neat, and to change the ratio of allyl alcohol to the diketene acetone from 1.5 eq to 1 eq allyl alcohol.23 In the next step, -keto ester 37 was reacted with acetaminobenzenesulfonyl azide to give the corresponding -diazo- -dicarbonyl compound as intermediate, which after addition of aq. LiOH and stirring at room temperature for 10 h gave diazoacetate 30.24 As an alternative to the expensive acetaminobenzenesulfonyl azide, freshly prepared p-toluenesulfonyl azide was used, but the yield obtained was only 50% on average and the reproducibility was poorly.

O

O

O+ OH

120 °C, 15 h

72%

46%

0.5 mol% Rh2(4R-MEOX)4CH2Cl2, reflux, 20 h

O

O

O

OH

DIBALH, CH2Cl275 °C, 1 h

OH t-BuOK THF, rt, 3 h

19PPh3Br

O

O O

O

ON2

Et3N, LiOH, CH2Cl2, rt, 18 h73%

35 36 37

302987% ee

38 39

SO

ON3N

H

O

Scheme 2.8. Synthesis of common building block lactone 30.

The cyclopropanation reaction turned out to be very difficult to perform, because it required a complex reaction set-up, in which diazoacetate 30 could be added slowly to a refluxing solution of the catalyst in dichloromethane. A solution to this problem was the use of a long cannula, which ended in the reaction flask passing the condenser. In this way lacton 29 was obtained with 46% yield. Unfortunately, even

Chapter 2

30

the smallest leakage in the reaction set-up allowed the volatile lactone to evaporate together with the solvent.

Our efforts to synthesize the end part of the meromycolate chain, alcohol 39, failed so far. The reduction to lactol 38 seems to occur, but there was an unidentified side product, which was difficult to separate by column chromatography. Even with the impurity, the lactol opening with the Wittig reagent was not observed. The significant NMR signals for the double bond were not detected.25

2.4.2 A model substrate of the -branched -hydroxy ester motif

In this study we synthesized a model substrate, testing the different steps to form the -hydroxy ester moiety as we envisioned for building block 19.26 The synthesis

described here was planned for a linear synthesis, in which first the meromycolate chain is connected to a -keto ester before the asymmetric hydrogenation is performed (Scheme 2.9).

[(RuCl(T-BINAP))2(μ-Cl)3NH2Me2]EtOH, 20 bar, 50 °C, 25 h

95%

OH

20

I2, PPh3, ImidazolTHF, 55 °C, overnight

72%

I

20

O

OH O

17 C22H45

DIPA, MeLi, HMPATHF, 40 °C, 7 h

40%

+O

O O NaH, n-BuLiTHF, 0 °C to rt, 3 h

O

O O

1759%

O

OH O

17

+

40 41 42

43>95% ee

44

45

Br

17

Scheme 2.9. Synthesis of the -branched -hydroxy ester motif.

Studies towards the synthesis of mycolic acid

31

The alkylation of -keto ester 41 with 1-bromooctadecane gave -keto ester 42. In the subsequent step an asymmetric hydrogenation was performed. First we followed the procedures described by Brückner et al. and Genet et al., using (R)-BINAP together with RuCl2(p-cymene), as the metal precursor, but no product formation was observed.27,28 Due to better results in other projects of our group, we changed the catalyst to the commercially available (R)-[(RuCl(Tol-BINAP))2( -Cl)3NH2Me2] complex. Complete conversion was subsequently observed and the isolated yield was high with 95%, but the ee determination was tricky since there is no UV-active group in the molecule. After protection of the secondary alcohol with a TBDPS group, still no separation on either chiral GC or chiral HPLC was found. As an alternative, we prepared the Mosher ester of the enantiopure compound and the racemate. In the end we observed for the enantioenriched compound only one signal in the 19F-NMR (compared to the racemate), which means that the ee is at least 95%. The long alkyl iodide 44 was prepared in one step from the corresponding alcohol. In a Fráter alkylation the mycolic chain 44 is introduced and a high diastereoselectivity is expected, due to the chiral induction of the hydroxyl group. Till now the reaction was only carried out with the racemate. The low yield of -hydroxy ester motif 45 is in accordance with earlier reported yields.17,26

2.5 Conclusion

The synthesis of an enantiopure -mycolic acid has not been completed, due to problems in the scaling-up of the cyclopropanation and the difficulties to perform the reaction with reproducible results. Sadder and wiser, for a future approach, meso diester 2 should be considered as common cyclopropyl building block, since this methodology of desymmetrization is well established and gives many options for connecting the alkyl chain linkers between the cyclopropyl units in the meromycolate chain. For the -branched -hydroxy ester motif, we applied a new commercially available ruthenium catalyst in the asymmetric hydrogenation of -keto esters with an C18 alkyl chain reaching high yields and an excellent enantioselectivity. Still it needs to be studied, whether similar results can be obtained with an additional functional group in the alkyl chain, since the synthetic route was changed. The Fráter alkylation needs to be investigated with an enantiopure compound, and the diastereoselectivity needs to be reviewed. The low yield can be avoided by first introducing an allyl group and elongation of the alkyl chain to the full length of the mycolic chain, as reported by Baird et al. in their latest synthesis.29

Chapter 2

32

2.6 Experimental

2.6.1 General information

All reactions were carried out in flame dried glassware under N2 atmosphere and using standard Schlenk technique. All dry solvents were taken from an MBraun solvent purification system (SPS-800). All chemicals were purchased from Acros, Sigma-Aldrich or TCI Europe, and used without further purification. Flash chromatography was performed using Screening Devices silica gel type SiliaFlash P60 (230 – 400 mesh). TLC analysis was performed on Merck silicagel 60/Kieselguhr F254, 0.25 mm and visualized by UV and staining with Seebach’s reagent. 1H and 13C NMR were recorded on a Varian 400-MR (400, 100.59 MHz, respectively) using CDCl3 as solvent. Chemical shift values are reported in ppm with the solvent resonance as the internal standard (CDCl3: 7.26 for 1H, 77.0 for 13C). Data are reported as follows: chemical shifts, multiplicity (s=singlet, d=doublet, t=triplet, q=quartet, br=broad, m=multiplet), coupling constants J (Hz), and integration. Carbon assignments are based on APT 13C NMR experiments. Progress of the reaction and conversion were determined by GC-MS (GC, HP6890: MS HP5973) with HP1 or HP5 columns (Agilent Technologies, Palo Alto, CA). Enantiomeric excess (ee value) was determined by chiral HPLC analysis using a Shimadzu LC-10ADVP HPLC equipped with a Shimadzu SPD-M10AVP diode array detector and chiral columns as indicated. Retention times (tR) and integrated ratios were obtained using Agilent Chemstation Software. High resolution mass spectra (HRMS) were recorded on a ThermoScientific LTQ Oribitrap XL spectrometer. Optical rotations were measured on a Schmidt + Haensch polarimeter (Polartronic MH8) with a 10 cm cell (concentration c given in g/100 mL).

Studies towards the synthesis of mycolic acid

33

2.6.2 Synthesis of the meromycolic chain

(S)-methyl 2-oxazolidine-4-carboxylate (33)30

Methyl L-serine (1.26 g, 8.09 mmol, 2 eq) was dissolved in 40 mL of CH2Cl2 and cooled to 0 °C with an ice bath. To the reaction mixture was added triethyl amine (3.4 mL, 24.2 mmol, 6 eq) within 3 min. The solution was stirred for 10 min before a solution of triphosgene (1.2 g,

4.04 mmol, 1 eq) in 10 mL of CH2Cl2 was slowly added over 30 min. The reaction was stirred at 0 °C for another 2 h. Then 60 mL of diethyl ether were added and the reaction mixture was cooled to -78 °C to precipitate all Et3NHCl. After 30 min, the reaction mixture was allowed to reach rt and then filtered. The filter cake was washed with diethyl ether. The filtrate was concentrated in vacuo and the crude product was purified by column chromatography (SiO2, ethyl acetate) to yield oxazolidine 33 as an oil (582 mg, 3.74 mmol, 69%). 1H NMR (400.0 MHz, CDCl3): ppm 6.85 (brs, 1H), 4.56 (dd, J=12.9, 7.4 Hz, 1H), 4.51–4.39 (m, 2H), 3.76 (s, 3H).13C NMR (100.6 MHz, CDCl3): ppm 170.7 (C), 159.2 (C), 66.6(CH2), 53.6 (CH), 52.8 (CH3). Optical rotation [ ]D = -1.27 (c = 1.22, EtOH).

Rh2(4S-MEOX)4 (31)22

A one-neck round-bottom flask charged with ligand 33 (300 mg, 2.07 mmol, 8.3 eq), Rh2(OAc)4 (110.1 mg, 0.25 mmol, 1 eq) and 150 mL chlorobenzene was fitted to a Soxhlet extractor. The extractor was equipped with a thimble containing 6 g of a mixture of Na2CO3 and sand (2:1). The reaction mixture was heated to reflux and stirred for 16 h. Next the solvent was

removed under reduced pressure and the blue solid was dissolved in a minimal volume of methanol. The excess of ligand was separated from catalyst 31 by column chromatography (CN Bound silica purchased from SiliCycle, methanol to methanol:acetonitrile 98:2). The red colored fractions were combined and concentrated in vacuo. The solid was then recrystallized from dry acetonitrile (1.0 mL per 100 mg) to give bright red crystals of Rh2(4S-MEOX)4 31 (68.4 mg, 0.138 mmol, 55%). Anal. Calcd. for C20H24N4O2Rh2: C, 30.71; H, 3.09; N, 7.16. Found: C, 32.68; H, 3.45; N, 9.27. The elementary analysis is not matching considering the tolerance value of ± 0.4.

O NH

O

COOMe

Rh Rh

N O

O N COOMe

HMeOOC

H

OH

MeOOC

O N COOMe

H

Chapter 2

34

Allyl 3-oxobutanoate (37)23

A suspension of diketene acetone adduct 35 (33 mL, 0.25 mol, 1 eq) in allyl alcohol (17 mL, 0.25 mol, 1 eq) was heated to 120 °C and stirred at this temperature for 15 h. After cooling to rt, the

product was purified by vacuum distillation. The main fraction was collected at 80 °C (0.3 mbar) to afford -keto ester 37 (25.6 g, 0.18 mol, 72%). 1H NMR (400.0 MHz, CDCl3): ppm 5.95-5.77 (m, 1H), 5.35-5.12 (m, 2H), 4.56 (d, J=6.9 Hz, 2H), 3.42 (s, 2H), 2.20 (s, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 200.2 (C), 166.5 (C), 131.4 (CH), 118.5 (CH2), 65.6 (CH2), 49.7 (CH2), 29.9 (CH3).

Allyl 2-diazoacetate (30)24

In 5 mL dry acetonitrile, -keto ester 37 (909 mg, 6.4 mmol, 1 eq) was dissolved and Et3N (1.1 mL, 8.32 mmol, 1.3 eq) was added. A solution of acetaminobenzensulfonyl azide (2.00 g, 8.32 mmol, 1.3

eq) in 5 mL acetonitrile was added dropwise in 5 min. This reaction mixture was stirred for 40 min at rt. Then aq. LiOH (805 mg, 19 mmol LiOH in 7 mL H2O) was added and the mixture was stirred for 10 h at rt. Next, the layers were separated and the water layer was extracted three times with a mixture of Et2O:EtOAc (2:1), the combined organic phases were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 9:1) to give diazoacetate 30 as bright yellow oil (590 mg, 4.67 mmol, 73%). 1H NMR (400 MHz, CDCl3 5.91-5.86 (m, 1H), 5.24 (dd, J=31.6, 13.8 Hz, 2H), 4.76 (br s, 1H), 4.61 (d, J=4.7 Hz, 2H). 13C NMR (100 MHz, CDCl3 170.8 (C), 132.4 (CH), 118.4 (CH2), 65.5 (CH2), 46.3 (CH). IR (film, cm)-1: 3122, 2948, 2887, 2111, 1691, 1445, 1383, 1181, 992, 836, 741.

(1R,5S)-3-oxabicyclo[3.1.0]hexan-2-one (29)19

To a refluxing solution of Rh2(4S-MEOX)4 (4 mg, 0.008 mmol, 0.5 mol%) in 10 mL CH2Cl2 was added diazoacetate 30 (200 mg, 1.58 mmol, 1 eq) in 10 mL CH2Cl2 over 10 h. After the addition, the reaction mixture was stirred for 15 h under reflux (precaution was taken that the reaction flask was

closed air tight; the product is volatile). The reaction mixture was cooled to rt and most of the CH2Cl2 was removed by distillation at atmospheric pressure. The catalyst was removed by column chromatography (SiO2, CH2Cl2) and gave lactone 29 as pale yellow oil (71 mg, 0.73 mmol, 46%, 87% ee). The enantiomeric ratio was determined by chiral GC analysis, Chiraldex -cyclodextrin column, 90 °C isotherm, retention time (min): 23.5 (major) and 24.3 (minor). 1H NMR (400 MHz, CDCl3 4.30-

O

O O

O

ON2

O

O

Studies towards the synthesis of mycolic acid

35

4.23 (m, 1H), 4.15-4.11 (m, 1H), 2.21-2.14 (m, 1H), 1.99-1.92 (m, 1H), 1.22-1.16 (m, 1H), 0.80-0.75 (m,1H). 13C NMR (100 MHz, CDCl3 176.3 (C), 69.3 (CH2), 17.3 (CH), 17.1 (CH), 12.0 (CH2).

2.6.3 Model substrate of the -branched -hydroxy ester motif26

Methyl 3-oxohenicosanoate (42)31

NaH (250 mg, 6.20 mmol, 1.5 eq) was suspended in 10 mL THF and cooled to 0 °C with an ice bath. At this temperature, -keto ester 41 (479 mg, 4.13 mmol, 1 eq) was added and the reaction mixture was

stirred for 10 min before n-BuLi (3.2 mL, 1.3 M in hexane, 1.1 eq) was added dropwise. After stirring the reaction mixture for 40 min at 0 °C, a solution of alkylbromide 40 (1.51 g, 4.54 mmol, 1.1 eq) in 5 mL THF was added dropwise. The reaction mixture was stirred for 3 h and during this time rt was reached, then the reaction was quenched by addition of aq. 2 M HCl. The layers were separated and the water layer was extracted three times with diethyl ether, the combined organic phases were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 9:1) to give -keto ester 42 as bright yellow oil (864 mg, 2.44 mmol, 59%). 1H NMR (400 MHz, CDCl3 3.74 (s, 3H), 3.44 (s, 2H), 2.54 (t, J=7.4 Hz, 2H), 1.64 (s, 2H), 1.26 (s, 30H), 0.88 (t, J=7.1 Hz, 3H). 13C NMR (100 MHz, CDCl3 203.0 (C), 167.8 (C), 52.5 (CH3), 49.2 (CH2), 43.3 (CH2), 32.2 (CH2), 29.9 (CH2), 29.8 (CH2), 29.7 (CH2), 29.6 (CH2), 29.2 (CH2), 23.7 (CH2), 22.9 (CH2), 14.3 (CH3).

(R)-methyl 3-hydroxyhenicosanoate (43)

-Keto ester 42 (177 mg, 0.50 mmol, 1 eq) was dissolved in 5 mL EtOH. Then (R)-(RuCl(T-BINAP))2( -Cl)3[NH2Me2] 5 mol% was added and the reaction mixture was stirred for 30 min at rt. The

reaction vial was placed in a stainless steel autoclave and first three times flushed with nitrogen, before a hydrogen pressure of 20 bar was applied. The reaction mixture was heated to 50 °C and stirred at this temperature for 72 h. After cooling to rt the autoclave was flushed with nitrogen. The crude product was filtered over silica and the column was washed with CH2Cl2. -Hydroxy ester 43 (169 mg, 0.47 mmol, 94%, 95% ee) was obtained after purification by column chromatography (SiO2, pentane/diethyl ether 3:2). The enantiomeric ratio was determined by Mosher ester analysis using 19F-NMR (racemate: -71.4, -71.5; enantiopure: -71.4).1H NMR (400 MHz, CDCl3 J=7.4 Hz, 2H), 1.64 (s, 2H), 1.26 (s, 30H), 0.88 (t, J=7.1 Hz, 3H). 13C NMR (100 MHz, CDCl3

O

O O

17

O

OH O

17

Chapter 2

36

(CH), 51.7 (CH3), 41.1 (CH2), 36.5 (CH2), 31.9 (CH2), 29.7 (CH2), 29.6 (2XCH2), 29.5 (2XCH2), 29.3 (CH2), 25.5 (CH2), 22.7 (CH2), 14.1 (CH3). HRMS was not measured.

1-iododocosane (44)32

In dry THF, iodine (2.79 g, 11.0 mmol, 1.1 eq), PPh3 (2.62, 10 mmol, 1 eq), imidazole (661 mg, 10 mmol, 1 eq) and docosan-1-ol (3.5 g, 11 mmol, 1.1 eq) were dissolved. This reaction mixture was stirred at 55 °C for 4 d.

After cooling, the reaction was quenched by addition of aq. NaHCO3. The layers were separated and the water layer was extracted three times with diethyl ether, the combined organic phases were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 9:1) to give 1-iododocosane 44 as light yellow solid (3.15 g, 7.23 mmol, 72%). 1H NMR (400 MHz, CDCl3 3.18 (t, J=6.0 Hz, 2H), 1.89-1.76 (m, 2H), 1.33 (s, 38H), 0.88 (t, J=7.6 Hz, 3H). 3C NMR (100 MHz, CDCl3

33.6 (CH2), 32.0 (CH2), 30.6 (CH2), 29.7 (CH2), 29.6 (2XCH2), 29.5 (CH2), 29.4 (CH2), 28.6 (CH2), 22.7 (CH2), 14.1 (CH3), 6.9 (CH2).

Methyl 2-(1-hydroxynonadecyl)tetracosanoate (45)

To a stirred solution of diisopropylamine (0.25 mL, 1.76 mmol, 3.3 eq) in 6 mL THF was added MeLi (1.4 M solution in hexane, 1.2 mL, 1.7 mmol, 3.2 eq) at -78 °C. After the addition, the solution was stirred for 30 min at 0 °C. Then the reaction mixture was cooled to

-40 °C and a solution of -hydroxy ester 43 (120 mg, 0.53 mmol, 1 eq) in 2 mL THF was added. After stirring for 30 min at the same temperature HMPA (0.58 mL, 3.34 mmol, 6.3 eq) and a solution of 1-iododocosane 44 (220 mg, 0.50 mmol, 1 eq) in 2 mL THF were added, and the reaction was stirred for 6 h at -40 °C. The reaction mixture was quenched by adding sat. aq. NH4Cl. The layers were separated and the water layer was extracted three times with diethyl ether, the combined organic phases were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 9:1) to give alkylation product 45 as light yellow solid (142 mg, 0.21 mmol, 40%). 1H NMR (400 MHz, CDCl3 -3.44 (m, 1H), 2.56-2.36 (m, 1H), 1.55-1.49 (m, 2H), 1.43 (m, 2H), 1.26 (s, 72H), 0.88 (t, 6H). 13C NMR (100 MHz, CDCl3 (CH3), 41.1 (2xCH2), 36.5 (2xCH2), 31.9 (2xCH2), 29.7 (2xCH2), 29.6 (2XCH2), 29.4 (2xCH2), 25.5 (CH2), 22.7 (CH2), 14.1 (CH3). HRMS was not measured.

I

20

O

OH O

17 C22H45

Studies towards the synthesis of mycolic acid

37

2.7 References

1. World Health Organisation, 2009, Global Tuberculosis Control, Geneva, Switzerland.

2. J. P. Cegielski, Clin. Infect. Dis., 2010, 50, S195–S200.

3. G. S. Besra and L. Kremer, Expert Opin. Investig. Drugs, 2002, 11, 1033–1049.

4. S. Vander Beken, J. R. Al Dulayymi, T. Naessens, G. Koza, M. Maza-Iglesias, R. Rowles, C. Theunissen, J. De Medts, E. Lanckacker, M. S. Baird, and J. Grooten, Eur. J. Immunol., 2011, 41, 450–460.

5. M. Daffé, J.-M. Reyrat, The mycobaterial cell envelope, AMS Press, Washington DC., 2008.

6. J. A. Verschoor, M. S. Baird, and J. Grooten, Prog. Lipid Res., 2012, 51, 325–339.

7. C. E. Barry III, R. E. Lee, K. Mdluli, A. E. Sampson, B. G. Schroeder, R. A. Slayden, and Y. Yuan, Prog. Lipid Res., 1998, 37, 143–179.

8. F. Laval, M.-A. Lanéelle, C. Déon, B. Monsarrat, and M. Daffé, Anal. Chem., 2001, 73, 4537–4544.

9. D. E. Minnikin and N. Polgar, Tetrahedron Lett., 1966, 7, 2643–2647.

10. M. Watanabe, Y. Aoyagi, M. Ridell, and D. E. Minnikin, Microbiology, 2001, 147, 1825–1837.

11. J. M. Viader-Salvadó, C. A. Molina-Torres, and M. Guerrero-Olazarán, J. Microbiol. Methods, 2007, 70, 479–483.

12. G. Shui, A. K. Bendt, I. A. Jappar, H. M. Lim, M. Laneelle, M. Hervé, L. E. Via, G. H. Chua, M. W. Bratschi, S. Z. Zainul Rahim, A. L. T. Michelle, S.-H. Hwang, J.-S. Lee, S.-Y. Eum, H.-K. Kwak, M. Daffé, V. Dartois, G. Michel, C. E. Barry, and M. R. Wenk, EMBO Mol. Med., 2012, 4, 27–37.

13. G. Koza and M. S. Baird, Tetrahedron Lett., 2007, 48, 2165–2169.

14. J. R. Al Dulayymi, M. S. Baird, E. Roberts, M. Deysel, and J. Verschoor, Tetrahedron, 2007, 63, 2571–2592.

15. J. R. Al Dulayymi, M. S. Baird, E. Roberts, and D. E. Minnikin, Tetrahedron, 2006, 62, 11867–11880.

16. J. R. Al-Dulayymi, M. S. Baird, H. Mohammed, E. Roberts, and W. Clegg, Tetrahedron, 2006, 62, 4851–4862.

Chapter 2

38

17. J. R. Al Dulayymi, M. S. Baird, and E. Roberts, Tetrahedron, 2005, 61, 11939–11951.

18. T. Niimi, T. Uchida, R. Irie, and T. Katsuki, Adv. Synth. Catal., 2001, 343, 79–88.

19. M. P. Doyle, R. E. Austin, A. S. Bailey, M. P. Dwyer, A. B. Dyatkin, A. V. Kalinin, M. M. Y. Kwan, S. Liras, and C. J. Oalmann, J. Am. Chem. Soc., 1995, 117, 5763–5775.

20. M. P. Doyle and W. Hu, J. Org. Chem., 2000, 65, 8839–8847.

21. M. P. Doyle, W. R. Winchester, M. N. Protopopova, P. Müller, G. Bernardinelli, D. Ene, and S. Motallebi, Helv. Chim. Acta, 1993, 76, 2227–2235.

22. M. P. Doyle, W. R. Winchester, J. A. A. Hoorn, V. Lynch, S. H. Simonsen, and R. Ghosh, J. Am. Chem. Soc., 1993, 115, 9968–9978.

23. G. Broggini, L. Garanti, G. Molteni, and G. Zecchi, Tetrahedron Asymmetry, 1999, 10, 487–492.

24. B. D’Abrosca, S. Pacifico, M. Scognamiglio, N. Tsafantakis, E. Pagliari, P. Monaco, and A. Fiorentino, Helv. Chim. Acta, 2013, 96, 1273–1280.

25. A.-Y. Park, H. R. Moon, K. R. Kim, M. W. Chun, and L. S. Jeong, Org. Biomol. Chem., 2006, 4, 4065–4067.

26. M. Nishizawa, H. Yamamoto, H. Imagawa, V. Barbier-Chassefière, E. Petit, I. Azuma, and D. Papy-Garcia, J. Org. Chem., 2007, 72, 1627–1633.

27. R. Kramer and R. Brückner, Angew. Chem. Int. Ed., 2007, 46, 6537–6541.

28. C. Roche, N. Desroy, M. Haddad, P. Phansavath, and J.-P. Genet, Org. Lett., 2008, 10, 3911–3914.

29. G. Toschi and M. S. Baird, Tetrahedron, 2006, 62, 3221–3227.

30. A. Fürstner, D. Kirk, M. D. B. Fenster, C. Aïssa, D. De Souza, C. Nevado, T. Tuttle, W. Thiel, and O. Müller, Chem. Eur. J., 2007, 13, 135–149.

31. G. Barbieri, G. Seoane, J.-L. Trabazo, A. Riva, F. Umpierrez, L. Radesca, R. Tubio, L. D. Kwart, and T. Hudlicky, J. Nat. Prod., 1987, 50, 646–649.

32. H. C. Huang, J. K. Rehmann, and G. R. Gray, J. Org. Chem., 1982, 47, 4018–4023.

Mimicking the natural cell membrane for mechanosensitive channels: synthesis of glycerophospholipids based on lactobacillic acid

In this chapter the synthesis of phosphatidylcholine lipids with mixed acyl chains is studied using a recently developed strategy. One part describes our effort to develop a new asymmetric approach to synthesize lactobacillic acid, a fatty acid possesing a cis-cyclopropyl moiety. To build the cis-configured cyclopropyl ring, we applied hetero asymmetric allylic alkylation and ring-closing metathesis as synthetic key steps. In a second part, the assembly of the acyl chains to a glycerol backbone and the introduction of the phosphatidylcholine headgroup are investigated.

Parts of this chapter have been published:

J.F. Teichert, T. den Hartog, M. Hanstein, C. Smit, B. ter Horst, V. Hernandez-Olmos, B.L. Feringa, A.J. Minnaard ACS Catalysis, 2011, 1, 309-315.

Chapter 3

40

3.1 Introduction

In each living organism, cells are the basic components, inheriting the machinery for proliferation. Cell membranes define the inside of the compartment from the external environment. The transport in and out of the cell is organized by membrane proteins.1 Notable examples of those membrane proteins are mechanosensitive proteins (MS), which are essential for the survival of the cell in case of osmotic downshift. The mechanosensitive proteins will detect tension in the cell wall and will respond to this with structural rearrangements forming non-selective pores to release inner pressure. In the field of sensory systems, one of the best-studied bacterial mechanosensors is the Mechanosensitive channel of Large conductance (MscL).2 MscL channels were first isolated from E. coli and further investigation showed that the channels could be reconstituted in artificial liposomes keeping their full function. Later, the structure of MscL channels was revealed by the X-ray structure of the MscL homologue from M. tuberculosis (Figure 3.1). It shows a channel of 85 Å length and two domains, which are organized as a homopentamer.3,4

Figure 3.1. Crystal structure of MscL from M. tuberculosis (PDB accession code 2OAR).

Most research today still focuses on the understanding of the mechanism of mechanosensation, specifically the influence of the channel gating depending on the protein lipid interaction. Most of the measurements are done in vitro in liposomes formed from commercially available glycerolphospholipids.5 This group of lipids can be divided in different subgroups depending on the phosphorus head group. The main lipids found in cell membranes contain phosphatidylcholine (PC) and phosphatidylethanolamine (PE) head groups with either saturated or unsaturated fatty acid chains or a mixture (Figure 3.2).1 In the cell membrane of Lactococcus lactis a special fatty acid bearing a cyclopropyl unit is found, called lactobacillic acid.6,7

Synthesis of a phosphoglycolipids based on lactobacillic acid

41

Considering the importance of the interactions between the MscL proteins and the lipids of the cell membrane, we tried to synthesize PC lipid 1 with lactobacillic acid as one of the fatty acid chains to mimic the natural membrane environment of Lactococcus lactis in liposomes. In addition we hoped to find, that by using a mixture of the commercially available PC/PE lipids and our synthesized glycerolphospholipid 1 the formed liposomes will be more stable.

phosphatidylcholine (PC)

phosphatidylethanolamine (PE)

O

OO

O

O

PO

OO

N

PC head grouplactobacillic acid

glycerolphospholipid 1

O

O

O

O

O

PO

ON

O

O

O

O

O

O

PO

ONH3

O

Figure 3.2. Chemical structure of PC and PE lipids and lactobacillic acid.

3.2 Reported syntheses of lactobacillic acid

Hofmann et al. were the first to isolate lactobacillic acid 2 from Lactobacillus arabinosus and Lactobacillus casei in 1950. The same group identified the structure of the fatty acid as a C18 chain with a cis-cyclopropyl moiety at position C11 and C12 according to degradation experiments.6,8,9 The absolute configuration was later established by Minnikin et al. by comparison of the natural product with the synthesized, enantiopure lactobacillic acid.10 Until today there have been only three enantioselective syntheses described. In the first published synthesis, an enzymatic desymmetrization was the key step to get an enantioenriched cyclopropyl ring.11 The second approach started with a building block derived from D-mannitol, e.g. from

Chapter 3

42

the chiral pool.10 In the latest synthesis the key step to build the cyclopropyl moiety was an asymmetric iridium-catalyzed cyclopropanation.12

3.2.1 Enzymatic desymmetrisation of meso cyclopropyl diester

In 1993, Kobayashi and co-workers were the first to use an enzyme catalyzed desymmetrization in the synthesis of a fatty acid bearing a cyclopropyl moiety in the chain (Scheme 3.1).11 In comparison to lactobacillic acid, the compound had a chain length of only C16, still this synthesis is considered here.

CO2MeMeO2CHH

1) PLE, pH 8.0 phosphate buffer, rt, 24 h2) BH3

.SMe2, B(OMe)3, THF, 20 °C to rt, 24 h3) p-TsOH, benzene, reflux, 30 min

O

HH

O

1) DIBAlH, CH2Cl2/hexane, 78 °C, 1 h2) Ph3P+C6H12CO2HBr-, NaCH2SOCH3 DMSO, rt, 1 h3) CH2N2, Et2O

HOH2C O

OPCC, NaOAc, MS 3 Å

CH2Cl2, rt, 1 h

OHC O

O

Ph3P+C5H11Br-, NaCH2SOCH3 DMSO, rt, 1 h

O

O

KO2CN=NCO2K, AcOH, MeOHreflux, 8 h

O

O

3 4

56

7

8overall yield 22%

90%

49%

70%

74%

95%

Scheme 3.1. Synthesis of lactobacillic acid by Kobayashi et al..

In the first step the meso diester 3 reacted in buffer with pig liver esterase to a monoester, which was immediately converted further to lactone 4. Next, 4 was reduced to a hemi-acetal, which then reacted with a Wittig reagent to ester 5. The free alcohol function of ester 5 was selectively oxidized with PCC to aldehyde 6, which

Synthesis of a phosphoglycolipids based on lactobacillic acid

43

reacted again with a Wittig reagent to elongate the carbon chain. The double bonds of the unsaturated ester 7 were reduced with in situ generated diimide to give the methyl ester of lactobacillic acid 8 in an overall yield of 22%.

3.2.2 Chiral pool strategy based on D-mannitol

Minnikin and co-workers published in 2003 an eight step synthesis based on D-glyceraldehyde 9 made out of D-mannitol (Scheme 3.2).10 The first step in the synthesis is a Wittig olefination of aldehyde 9 leading to acetal 10. In the following step the cyclopropyl moiety is introduced by a Simmons-Smith reaction.

O

O

O

Br-Ph3P+(CH2)6Men-BuLi, THF, 78 °C

82%

Et2Zn, ClCH2I DCE, 30 °C

90%

1) HCl, MeOH2) NaIO4, H2O

O

Me(H2C)4

HH

46%

Br-Ph3P+(CH2)9OTHPNaHMDS, THF, 78 °C

Me(H2C)4

HHTHPO(H2C)8

Me(H2C)4

HHHO(H2C)8+

1433%

1311%

N2H4, H2O, NaIO4 CH3COOH, CuSO4, i-PrOH

Me(H2C)4

HH Me(H2C)4

HH+

1658%

1566%

HO(H2C)9THPO(H2C)9

80%

p-TsA, CH2Cl2

RuO2, NaIO4 MeCN, MeOH, H2O

overall yield 1%

HHO

O

Me(H2C)4

HH

OO

Me(H2C)4

OH

2

9 10 11

12

20% 5 7

O

Scheme 3.2. Synthesis of lactobacillic acid by Minnikin et al..

In this reaction, the stereoselectivity is induced by the stereocenter in the acetal. The zinc carbenoid coordinates to the oxygen of the acetal and the methylene transfer is

Chapter 3

44

directed to the bottom face of the double bond. Aldehyde 12 was formed after deprotection of 11 and periodate cleavage of the diol. The reaction of formylcyclopropane 12 with a Wittig reagent possessing a THP protected alcohol gave two products, the protected alcohol 13 and the already deprotected alcohol 14. Synthesis was continued with the mixture and the double bond originating from the Wittig reaction was reduced with diimide to give products 15 and 16. Then alcohol 15 was deprotected and the combined amounts of alcohol 16 were oxidized with RuO2 and NaIO4 to give lactobacillic acid 2 with an overall yield of 1%.

3.2.3 Asymmetric iridium-catalyzed cyclopropanation

The latest synthesis is from 2008 reported by Katsuki and co-workers. In this approach for the first time asymmetric transition metal-catalysis is used to build the cyclopropyl moiety enantioselectively (Scheme 3.3).12 The synthesis begins with the key step, the formation of the cyclopropyl ring by cyclopropanation. 1-Octene reacted with an -diazoacetate in the presence of iridium-salen catalyst 19 to give ester 20 in 60% yield and an excellent enantioselectivity of 98%. In the following operations the complete chain of lactobacillic acid was build. Intermediate 21 was formed by reduction of ethyl ester 20 with LiAlH4. Then the alcohol was converted into sulfone 22 via Mitsunobu reaction with 1-tert-butyltetrazole-5-thiol and oxidation. In a Julia-Kocienski olefination of sulfone 22 with the appropriate aldehyde the full length of the carbon chain of lactobacillic acid was formed. In the last step, the double bond next to the cyclopropyl moiety was reduced with diimide under mild conditions to give the methyl ester of lactobacillic acid 8. Overall, the number of steps could be reduced compared to the earlier reported syntheses by Kobayashi and Minnikin, and the overall yield increased to 39%.

Albeit in this synthesis the cyclopropyl moiety is synthesized by an asymmetric catalytic approach, the drawback is the seven step synthesis of the iridium-salen complex 19.

Synthesis of a phosphoglycolipids based on lactobacillic acid

45

1 mol% 19 THF, 78 °C, 48 h

+ EtO

O

N2

CO2Et68%98% ee

LiAlH4, THF0 °C to rt, 2 h

OH

99%

1) PPh3,DEAD, Ph-tetrazole-SH THF, 0 °C to rt, 3 h 2) H2O2,EtOH, H2O, (NH4)6Mo7O24

.4H2O, 24 hSO

ON N

NN

Ph90%

LiHMDS, THF, 78 °CMeO2C(CH2)5CHO, 78 °C to rt

73%

OMe

O

KO2CN=CNCO2K, AcOHMeOH, reflux, 8h 88%

OMe

O

overall yield 39%

17 18 20

2122

23

8

N N

PhPhO O

IrL

(aR,R)-Ir salen 19

Scheme 3.3. Synthesis of lactobacillic acid by Katsuki et al..

3.3 Retrosynthetic analysis

Our first retrosynthetic step in the synthesis of glycerolphospholipid 1 is the cleavage of the phosphorus head group to give the glycerol part with two fatty acid residues (Scheme 3.4). Compound 25 can be disconnected into three building blocks: lactobacillic acid 2, commercially available (R)-TBDPS-glycidyl ether 26 and commercially available stearic acid 27.

Chapter 3

46

O OTBDMS

OP

ClO

O

OH

O

OH

O

O

O

O

O

O

O

O

O

PO

ON

OH

OH

O

+

+

+

1

2425

2 26

27

Scheme 3.4. Retrosynthetic analysis of PC lipid 1.

We envisioned for the synthesis of lactobacillic acid methyl ester 8 an asymmetric approach with lactone 28 as intermediate (Scheme 3.5), since the cis-configured double bond has a key function in the installation of the cis-cyclopropyl moiety. The double bond of lactone 28 was formed in a ring-closing metathesis (RCM) of a diene ester, assembled out of carboxylic acid 29 and chiral secondary alcohol 30. The secondary alcohol derives from the deprotection of allylic ester 31. The chirality in ester 31 was introduced in a hetero asymmetric allylic alkylation (h-AAA), which was established in our institute some years ago by the group of Feringa.13,14

Synthesis of a phosphoglycolipids based on lactobacillic acid

47

OMe

O

OO

O

O

OH

OH

O+

8

28

29 30

31

Scheme 3.5. Retrosynthetic analysis of lactobacillic acid methyl ester 8.

3.4 Results and discussion

3.4.1 Synthesis of lactobacillic acid

We saw a challenge in synthesizing lactobacillic acid in an asymmetric fashion and avoiding the synthesis of the complex salen ligand used in the synthesis of Katsuki. An alternatively route would be to build a chiral allylic alcohol with a cis-double bond and to introduce the cyclopropyl ring in a Simmons-Smith reaction, preferably with high stereoselectivity, due to pre-coordination of the carbenoid to the hydroxyl group. One option to synthesize chiral allylic alcohols is the copper-catalyzed h-AAA with Grignard reagents (Scheme 3.6).13

In the first step, ester 34 was synthesized from benzoyl bromide 32 and acrolein 33, which is the starting material for the h-AAA. In the h-AAA the earlier reported optimized conditions for substrate 34 were applied to get the chiral allylic alcohol 31 with 91% yield and 96% ee. Next, the alcohol was hydrolyzed with potassium

Chapter 3

48

hydroxide to give allylic alchohol 30. Carboxylic acid 29 was made in one step from 11-bromoundec-1-ene via its Grignard reagent and CO2 in the form of dry ice. The diene 35 was assembled from 30 and 29 in a Steglich esterification. Diene 35 was used subsequently as substrate for the RCM to give key intermediate 28.

O

Br + OCH2Cl2, rt, 72 h

36%

O

O Br

5 mol% L1 5 mol% CuBr.SMe2

2 eq C5H11MgBr 75 °C, CH2Cl2, 20 h

O

O

91%97% ee

OH

OH

O

O

O

OO

5 eq KOHMeOH, THF, H2O

rt, 20 h80%

DCC, DMAPCH2Cl2, rt, 16 h

91%

+

Grubbs II, toluene120 °C, 5 h

68%

32 33 34

31

30

29

35

28

FePPh2

N H

(S,Rfe) Taniaphos L1

PPh2 N NMe

Me Me

Me

Me Me

RuCl

Cl

PCy3

Grubbs II 36

Scheme 3.6. Synthesis of key intermediate 28 using a h-AAA to form an allylic ester.

The formation of medium-sized rings by RCM is still a challenge, especially in the absence of elements to control the conformation in the substrate. In the total synthesis

Synthesis of a phosphoglycolipids based on lactobacillic acid

49

of herbarumin I and II, Fürstner et al. described the possibility to control the configuration of the double bond in a ten-membered lactone by choosing different catalysts for the reaction. The kinetic product (E-isomer) was selectively formed using Grubbs first generation catalyst and on the other hand performing the RCM with Grubbs second generation catalyst the thermodynamic product (Z-isomer) was observed as the only product.15 Therefore we chose for our reaction the Grubbs second generation catalyst to selectively build the cis-configured double bond of lactone 28.

Fürstner explains the selectivity for the Z-isomer with the higher activity of the second generation catalyst in comparison to the one of the first generation and the related ability to isomerize the double bond to the thermodynamically favored product over time.16,17 First we performed the reaction in dichloromethane and observed a mixture of E/Z-isomers. We followed the formation of the Z-isomer from the E-isomer by NMR to identify the time, in which the isomerization is finished. In the 1H-spectrum, the signal of proton 1 of the double bond increases over time and the formation of the Z-isomer is clearly visible. The reaction seems to stop after 15 h and not all E-isomer is converted, this might be due to deactivation of the catalyst (Figure 3.3), therefore an E/Z mixture of 1:3 was observed. Later, the same reaction was investigated in toluene. Here, the reaction is much faster and the formation of the E-isomer was not observed.

Figure 3.3. NMR-study of the formation of the cis-configured lactone 28.

Chapter 3

50

After isolation and purification we wanted to show that lactone 28 is cis-configured, this was established by NMR studies. From the 1H-spectrum the coupling constant between the two protons of the double bond indicates already a cis-configured double bond. For additional proof a NOESY experiment was performed, which shows only cross-couplings between protons through space (Figure 3.4).

Figure 3.4. Cross-peaks between H1 and H2 in the NOESY spectrum of 28.

In the NOESY-spectrum two cross peaks are found, one between the proton at position 3 and the proton of the double bond at position 2. The second cross peak is between the two protons of the double bond at position 1 and 2, this established a cis relation in the carbon-carbon double bond.

The next step in the synthesis was the lactone opening to ester 36 (Scheme 3.7). With ester 36 in hand, the introduction of the cyclopropyl moiety by a Simmons-Smith reaction was performed. To find the optimal conditions, time was spent to investigate a model substrate (see paragraph 3.4.2). The reaction with ester 36 as substrate was carried out at -40 °C and allowed to reach -25 °C. After 4 h, the conversion was incomplete and more zinc reagent was added at -25 °C. After 16 h stirring at -25 °C most of the double bond had reacted and the reaction was worked up and in the

Synthesis of a phosphoglycolipids based on lactobacillic acid

51

1H-spectrum the two characteristic signals below 0.5 ppm for the two protons of the cyclopropyl ring were observed.

OO

OMe

O

OH

OMe

O

OH

OMe

O

MeOH, K2CO3

90%

a) Burgess reagentb) TsCl, basec) MsCl, based) DIAD, PPh3, DBU

SO

ONEt3N CO2Me

flavin, NH2NH2 O2, EtOH

OMe

O

Et2Zn, CH2I2, CH2Cl2 10 °C to rt, 12 h

NNH

O

N

OAc

OAcOAc

AcO

HN O

28 36

37

38

8

O NN O

O

O

Burgess reagent

DIAD

N-ethyl-riboflavin 39

71%

Scheme 3.7. Synthetic approach from intermediate 28 to lactobacillic ester 8.

In the following step the secondary alcohol had to be eliminated to give unsaturated ester 38. In the first attempt Burgess-reagent was used, which is normally used in dehydration reactions of sensitive compounds. In all reactions, varying the temperature no product formation was observed, but degradation instead. The formation of the tosylate or mesylate of the alcohol group completely failed no matter, which base was employed (pyridine, DBU, Et3N). As a last option Mitsunobu conditions were applied, first without base and later with DBU as the base. When DBU was added, finally some reaction was observed, but at the same time an

Chapter 3

52

unidentified by-product was observed as major product. The common problem in all the attempts to eliminate the hydroxyl group is most probably the formation of a carbocation. As soon as this is formed, a rearrangement of the ring is likely to occur. As a result of all failed attempts to eliminate the hydroxyl group next to the cyclopropyl ring, this approach to synthesize lactobacillic acid was terminated.

In the meantime, the flavin catalyzed diimide reduction to give ester 8 was successfully tested on a model substrate (see paragraph 3.4.2).

3.4.2 Model substrate for flavin reduction

The reduction of double bonds by diimide, which is generated in situ from hydrazine by the organocatalyst N-ethyl-riboflavin 39, was developed in our lab.18 We planned to apply this methodology to reduce the double bond in substrate 38. To ensure this gives the desired product, we synthesized ester 43 as test substrate by Simmons-Smith reaction of dienol 40 (Scheme 3.8).19 The reaction gave a mixture of alcohols 41 and 42. For an easier separation and purification of alcohol 41 the pivalic ester was formed to give the desired model substrate. Vinylcyclopropane 43 was successfully converted into the corresponding saturated cyclopropane 44 by the improved reduction method described earlier. The product was obtained in 84% yield without any side-reaction observed. This result is comparable to the yields reported for other examples of diimide reduction of vinyl cyclopropanes (see the described syntheses in part 3.2).

OH

Et2Zn, CH2I2,CH2Cl20°C to rt, 34 h

OH OH+88%

DCC, DMAP, CH2Cl2, 0 °C, 16 h

O

O

HO

O

58%

flavin 39, NH2NH2 EtOH, rt

O

O

84%

40 41 42

4344

Scheme 3.8. Organocatalytic reduction of vinylcyclopropane 42.

Synthesis of a phosphoglycolipids based on lactobacillic acid

53

3.4.3 A model substrate for the desired gycerolphospholipid

In the meantime while we tried to synthesize lactobacillic acid, we tested a new synthetic route for PC lipids, with the commercially available fatty acids oleic acid 45 and stearic acid (see Scheme 3.9). The enantioselective opening of glycidyl ether 26, using fatty acids as nucleophile, was established in our group for the synthesis of a PE lipid based on tuberculostearic acid.

In our synthesis of PC lipids, the key step is an enantioselective opening of a terminal epoxide. Jacobsen and co-workers reported in 1997 the opening of meso epoxides with carboxylic acids in the presence of cobalt-(salen) catalyst 46 giving products with high yield and excellent enantioselectivities. Later they showed that a kinetic resolution of terminal epoxides with the same catalyst and water as nucleophile gives high ee’s.20,21 Based on those results we envisioned to open a chiral enantiopure terminal epoxide with a fatty acid using catalyst 46.22

The first step was a cobalt-(salen)-catalyzed opening of epoxide 26 with oleic acid 45 to obtain chiral secondary alcohol 47. Without further purification, the crude product was converted with stearic acid, DCC and DMAP to diester 48. In the following step the TBDMS-group was deprotected by BF3 etherate in acetonitrile to result in compound 49, which was immediately used in the next reaction to avoid acyl chain migration to the free alcohol group. In two steps the PC head group was installed following the procedure by Acuña et al..23 First the free alcohol 49 reacted with 2-chloro-2-oxo-1,3,2-dioxaphospholane in the presence of Et3N to form a phospholane as intermediate. The last step was the ring opening of the phospholane by Me3N as gas under reflux in a pressure tube to give the final PC lipid 50 in 48% yield over two steps.

Chapter 3

54

O OTBDMS

NN

OO

tBu

tBu tBu

tBu

Co

1 mol% 46, DIPEAEt2O, rt, 2d

stearic acid, DCC, DMAPheptane, rt, over night

BF3.Et2O, MeCN, 2 h, rt

OP

ClO

O1) Et3N, THF, rt, 3 h

2) Me3N, MeCN, 70 °C, 4h

69% over two steps

91%

48%

26

+

4547

48

49

50

O

O

O

O

O

PO

ON

O

O OTBDMSOH

OTBDMS

O

O

O

O

OH

O

O

O

O

Co-(salen)46

oleic acid

O

Scheme 3.9. Synthesis of PC model lipid 50.

Synthesis of a phosphoglycolipids based on lactobacillic acid

55

3.5 Conclusion

Albeit we have not yet completely synthesized PC lipid 1 due to the unfinished synthesis of lactobacillic acid, we found a new way to build chiral internal allylic alcohols with a cis-configured double bond by performing a sequence of reactions, in which the h-AAA and the RCM are key steps. An additional achievement is the application of riboflavin catalyzed diimide reduction of vinylcyclopropyl compounds like substrate 43, showing the strength of this methodology.

For the synthesis of glycerophospholipid bearing a PC head group we were able to use cobalt-(salen) catalyst 46 established by Jacobsen et al. to open a terminal epoxide with a fatty acid as nucleophile and combine this with the installation of the PC head group via a phospholane intermediate.

Still one problem stayed unsolved: the dehydration of the hydroxy group next to a cyclopropyl moiety without rearrangement of the ring.

3.6 Experimental For general information see experimental of chapter 2.

3.6.1 Synthesis of lactobacillic acid

(E)-3-bromoprop-1-en-1yl benzoate (34)13

Freshly distilled acrolein (11.2 g, 200 mmol, 1.2 eq) was dissolved in 80 mL CH2Cl2 and cooled with an ice bath. Benzoyl bromide (30.9 g, 167 mmol, 1 eq) was added

dropwise and the reaction mixture was allowed to reach rt. After stirring for 3 d at rt the reaction was quenched by addition of saturated aq. NaHCO3 solution. The phases were separated and the water layer was extracted three times with CH2Cl2. The combined organic layers were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The residue was then diluted with 10 mL n-pentane (analytical quality) and stored at -20 °C for 6 h. The white crystals were filtered over a porcelain filter (suction filtration) and washed with cold n-pentane (crystals are quite soluble). To increase the yield the process was repeated. Allyl bromide 34 was isolated as a white solid (11.17 g, 46.3 mmol, 26%). 1H NMR (400.0 MHz, CDCl3): 8.08-8.11 (m, 2H), 7.67 (dt, J = 12.3, 1.0, 1H), 7.62 (m, 1H), 7.46-7.50 (m, 2H), 5.90 (dt, J = 12.3, 8.4, 1H), 4.07 (dd, J = 8.4, 1.0, 2H). 13C NMR (100.6 MHz, CDCl3): ppm 163.2 (C), 139.4

O

O Br

Chapter 3

56

(CH), 133.9 (CH), 130.1 (2xCH), 128.6 (2xCH), 128.4 (C), 111.8 (CH), 28.6 (CH2). Anal. calcd. for C10H9BrO2: C, 49.82; H, 3.76; Br, 33.14; O, 13.27. Found: C, 50.18; H, 3.78.

(S)-oct-1-en-3-yl benzoate (31)13

The Grignard reagent of 1-bromopentane was prepared by activation of magnesium (2.81 g, 115 mmol, 2.5 eq) with iodine and addition of 10 mL diethyl ether. To this was added 1-bromopentane (14.0 g, 92.6 mmol, 2.0 eq) in 20 mL

diethyl ether, dropwise to keep the reaction mixture refluxing. The reaction mixture was stirred for 18 h at rt. In a Schlenk tube allyl bromide 34 (11.2 g, 46.3 mmol, 1 eq) was dissolved in 200 mL CH2Cl2 and cooled to -76 °C. To this mixture was added

2 (58.9 mg, 0.286 mmol, 0.6 mol%) and (S,Rfe) Taniaphos (300 mg, 0.436 mmol, 0.9 mol%) and the mixture was stirred for 20 min at -76 °C. Then the Grignard reagent was added over 14 h and the mixture was stirred at -76 °C for another 6 h. The reaction was quenched with 50 mL MeOH and allowed to reach rt before 100 mL of saturated aq. NH4Cl solution was added. The layers were separated and the water layer was extracted three times with CH2Cl2. The combined organic layers were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 19:1 to 9:1) to yield allyl ester 31 as a colorless oil (9.84 g, 42.4 mmol, 91%, 96% ee). 1H NMR (400.0 MHz, CDCl3): ppm 8.09-8.06 (m, 2H), 7.57-7.53 (m, 1H), 7.46-7.42 (m, 2H), 5.90 (ddd, J = 6.3, 10.5, 17.0, 1H), 5.53-5.48 (m, 1H), 5.33 (dd, J = 1.3, 17.2, 1H), 5.20 (dd, J = 1.2, 10.5, 1H), 1.85-1.67 (m, 2H), 1.46-1.27 (m, 6H), 0.89 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 165.9 (C), 132.8 (CH), 130.6 (C), 129.6 (2x CH), 128.3 (2x CH), 116.5 (CH), 75.4 (CH), 34.3 (CH2), 31.6 (CH2), 24.8 (CH2), 22.8 (CH2), 14.0 (CH3). Optical rotation [ ]D = +29.9 (c = 1.01, CHCl3). HRMS (ESI+): m/z [M+Na] calcd for C15H20O2Na: 255.1356; found: 255.1364. Enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OB-H column, n-heptane/i-PrOH 99.5:0.5, 40 °C isotherm, detection at 254 nm, retention time (min): 7.05 (major) and 8.53 (minor).

(S)-oct-1-en-3-ol (30)24

Allyl ester 31 (9.80 g, 42.2 mmol, 1 eq) was dissolved in 120 mL of THF, MeOH, H2O (1:1:1) and KOH powder (23.7 g, 422 mmol, 10

eq) was added at rt. The reaction was stirred at rt for 17 h, and then water and diethyl ether were added. The phases were separated and the water layer was extracted five times with diethyl ether. The combined organic layers were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 3:1 to 1:1) to yield allyl alcohol

O

O

OH

Synthesis of a phosphoglycolipids based on lactobacillic acid

57

30 as colorless oil (4.87 g, 38.0 mmol, 80%). 1H NMR (400.0 MHz, CDCl3): ppm 5.89-5.81 (m, 1H), 5.20 (dd, J = 15.9, 1.3, 1H), 5.08 (dd, J = 9.3, 1.1, 1H), 4.07 (m, 1H), 1.50 (m, 2H) 1.35 (br, 6H), 0.88 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 141.3 (CH), 114.4 (CH), 73.2 (CH), 37.0 (CH2), 31.7 (CH2), 25.0 (CH2), 22.6 (CH2), 14.0 (CH3). Optical rotation [ ]D = +11.6 (c = 1.05, CHCl3). HRMS (ESI+): m/z [M-H2O]+ calcd for C8H15: 111.1168; found: 111.1164.

Dodec-11-enoic acid (29)25

For the preparation of the Grignard reagent, Mg (2.71 g, 111 mmol, 1.3 eq) was activated by an iodine crystal and stirred in 10 mL of diethyl ether. To this,

11-bromoundec-1-ene (20.0 g, 86 mmol, 1 eq) in 40 mL diethyl ether was added dropwise so that the reaction mixture was refluxing. After the addition was complete the reaction was stirred at 40 °C for 4 h. In a three-necked flask equipped with condenser and addition funnel, dry ice (37 g, 858 mmol, 10 eq) was added and the Grignard reagent was slowly added at rt. The reaction mixture was allowed to reach rt and stirred for 18 h. The reaction was quenched by addition of aq. 10% HCl solution. The phases were separated and the water layer was extracted three times with diethyl ether. The solvent was evaporated and the crude product was dissolved in 50 mL aq. 1 M NaOH. The solution was extracted three times with diethyl ether, combined organic layers were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 9:1 to 3:1) to yield acid 29 as a yellow oil (10.3 g, 52.1 mmol, 60%). 1H NMR (400.0 MHz, CDCl3): 6.29 (m, 1H), 5.44 (m, 2H), 2.83 (t, 2H), 2.52 (m, 2H), 1.76 (m, 12H). 13C NMR (100.6 MHz, CDCl3): ppm 180.3 (C), 139.2 (CH), 114.1 (CH2), 34.1 (CH2), 33.8 (CH2), 29.4 (2x CH2), 29.2 (CH2), 29.1 (CH2), 29.0 (CH2), 28.9 (CH2), 24.7 (CH2). HRMS (ESI+): m/z [M+Na]+ calcd for C12H21O2: 197.1536; found: 197.1538.

(S)-oct-1-en-3-yl dodec-11-noate (35)

Acid 29 (9.41 g, 47.5 mmol, 1.5 eq) was dissolved in 30 mL CH2Cl2 and cooled with an ice bath. DCC (13.0 g, 63.3 mmol,

2 eq) and DMAP (386 mg, 3.16 mmol, 0.1 eq) were added and stirred for 10 min. To this mixture a solution of alcohol 30 (4.06 g, 31.6 mmol, 1 eq) in 10 mL CH2Cl2 was added. The reaction was allowed to reach rt and stirred for 24 h. Then 100 mL ice cold n-pentane was added to the reaction mixture, afterwards the solution was filtered over celite. The filtrate was concentrated in vacuo and the crude product was

OH

O

O

O

Chapter 3

58

purified by column chromatography (SiO2, pentane/diethyl ether 50:1 to 10:1) to yield diene 35 as yellow oil (9.82 g, 31.8 mmol, 91%). 1H NMR (400.0 MHz, CDCl3): ppm 5.87-5.71 (m, 2H), 5.28–5.18 (m, 2H), 5.13 (d, J=6.8 Hz, 1H), 5.02–4.88 (m, 2H), 2.30 (t, 2H), 2.08–1.97 (m, 2H), 1.70–1.50 (m, 5H), 1.43–1.20 (m, 23H), 0.88 (t, J=6.9 Hz, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 173.1 (C), 139.2 (CH), 136.8 (CH), 116.3 (CH2), 114.1 (CH2), 74.5 (CH), 34.6 (CH2), 34.2 (CH2), 33.8 (CH2), 31.5 (CH2), 29.4 (2 x CH2), 29.2 (CH2), 29.1 (2 x CH2), 28.9 (CH2), 25.0 (CH2), 24.7 (CH2), 22.5 (CH2), 14.0 (CH3). Optical rotation [ ]D = -3.36 (c = 1.13, CHCl3). HRMS (ESI+): m/z [M+H]+ calcd for C20H37O2: 309.2788; found: 309.2784.

(S,Z)-14-pentyloxacyclotetradec-12-en-2-one (28)15

Diene 35 (600 mg, 1.94 mmol, 1 eq) was dissolved in 120 mL toluene and degassed by a nitrogen flow for 30 min. After the addition of Grubbs second generation catalyst (83 mg, 5 mol%), the reaction mixture was stirred at reflux for 5 h and the progress of the reaction was followed by NMR. The

reaction mixture was cooled to rt, filtered of a silica column and the column was flushed with CH2Cl2. The filtrate was concentrated in vacuo and the crude product was purified by column chromatography (SiO2, pentane/diethyl ether 90:10) to yield lactone 28 as yellow oil (369 mg, 1.32 mmol, 68%). 1H NMR (400 MHz, CDCl3

5.69-5.61 (m, 1H), 5.41–5.35 (m, 1H), 5.30-5.24 (m, 1H), 2.33–2.25 (m, 2H), 2.16–2.14 (m, 1H), 2.04-1.94 (m, 1H), 1.62 (m, 4H), 1.29 (br, 18H), 0.88 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 173.3 (C), 135.5 (CH), 130.0 (CH), 74.8 (CH), 34.1 (CH2), 33.8 (CH2), 31.5(CH2), 30.7 (CH2), 26.8(CH2), 25.9 (CH2), 25.6 (CH2), 25.2 (CH2), 25.0 (CH2), 24.9 (CH2), 24.5 (CH2), 22.5 (CH2), 22.3 (CH2), 14.0 (CH3). Optical rotation [ ]D = -33.3 (c = 1.04, CHCl3). HRMS (ESI+): m/z [M+Na] calcd for C18H32O2Na: 303.2295; found: 303.2303.

(S,Z)-methyl 13-hydroxyoctadec-11-enoate (36)24

Lactone 28 (164 mg, 0.59 mmol, 1 eq) was dissolved in 20 mL MeOH and K2CO3 (323 mg, 2.34 mmol, 4

eq) was added. The mixture was stirred for 24 h at rt. On TLC still some starting material was observed, therefore another portion of K2CO3 (323 mg, 2.34 mmol, 4 eq) was added and the reaction mixture was stirred at rt for 18 h. The reaction mixture was quenched by adding aq. NH4Cl. The phases were separated and the water layer was extracted three times with CH2Cl2, the combined organic phases were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The crude product was

OO

OMe

O

OH

Synthesis of a phosphoglycolipids based on lactobacillic acid

59

purified by column chromatography (SiO2, pentane/diethyl ether 3:1 to 1:1) to give carboxylic ester 36 as colorless oil (170 mg, 0.54 mmol, 90%). 1H MR (400 MHz, CDCl3 –5.58 (m, 1H), 5.44 (dd, J=15.3 Hz, 7.1, 1H), 4.03 (q, J=6.6 Hz, 1H), 3.66 (s, 3H), 2.30 (t, 2H), 2.01 (q, 2H), 1.63-1.53 (m, 5H), 1.28 (br s, 18H), 0.88 (t, 3H). 13C NMR (101 MHz, CDCl3 (CH), 51.4 (CH3), 37.3 (CH2), 34.1 (CH2), 32.1 (CH2), 31.7 (CH2), 29.4 (CH2), 29.3 (CH2), 29.2 (CH2), 29.1 (CH2), 29.1 (CH2), 29.0 (CH2), 25.2 (CH2), 24.9 (CH2), 22.6 (CH2), 14.0 (CH3). Optical rotation [ ]D = -2.0 (c = 1.20, CHCl3). HRMS was not measured.

methyl 10-((1R,2S)-2-((S)-1-hydroxyhexyl)cyclopropyl)decanoate (37)26

Ester 36 (119 mg, 0.38 mmol, 1 eq) was dissolved in 10 mL CH2Cl2 and cooled to -45 °C. To this solution

was added Et2Zn (1.5 mL, 1.52 mmol, 4 eq, 1 M in hexane) and CH2I2 (0.3 mL, 3.72 mmol, 9.5 eq) simultaneously with a syringe pump over 2 h. After the addition the reaction mixture was stirred for 5 h at -45 °C and further stirred for 2 h at -20 °C. Since the TLC control showed incomplete conversion, more Et2Zn (7 mL, 7 mmol, 20 eq) and CH2I2 (3.0 mL, 37.2 mmol, 98 eq) were added at -20 °C and the reaction mixture was stirred at this temperature for 18 h. Then the reaction mixture was quenched with aq. 3M NaOH. The phases were separated and the water layer was extracted three times with diethyl ether, the combined organic phases were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 3:1) to give cyclopropyl ester 37 as a yellow oil (92 mg, 0.27 mmol, 71%). 1H MR (400 MHz, CDCl3 J=7.6 Hz, 2H), 1.63-1.55 (m, 4H), 1.25 (br s, 20H), 0.88 (t, J=6.7 Hz, 3H), 0.67-0.53 (m, 2H), 0.46–0.36 (m, 1H), 0.28-0.23 (m, 1H). 13C NMR (100 MHz, CDCl3 CH3), 37.2 (CH2), 34.1 (CH2), 33.6 (CH2), 31.9 (CH2), 29.5 (CH2), 29.4 (CH2), 29.4 (CH2), 29.2 (CH2), 29.1(CH2), 25.8 (CH), 25.3 (CH2), 24.9 (CH2), 22.6 (CH2), 17.0 (CH), 14.0 (CH3), 9.81 (CH2). Optical rotation [ ]D = +10.7 (c = 1.03, CHCl3). HRMS (ESI+): m/z [M+Na] calcd for C20H38O3Na: 349.2713; found: 349.2711.

OMe

O

OH

Chapter 3

60

3.5.2 Model substrate for flavin reduction

(E)-(2-(prop-1-en-1yl)cyclopropyl)methanol (41)/(42)27

A solution of diethylzinc (25.5 mL, 25.5 mmol, 2.5 eq, 1 M in hexane) in CH2Cl2 (40

mL) was cooled to 0 °C. Diiodomethane (2.5 mL, 30.6 mmol, 3.0 eq) was added dropwise. The resulting mixture was stirred at the same temperature for 15 min before 2,4-hexadien-1-ol 39 (1.2 mL, 10.2 mmol, 1 eq) was added dropwise. The reaction mixture was allowed to reach rt. The progress of the reaction was followed by GC-MS. After 30 h, diethylzinc (10 mL, 10 mmol) was added at 0 °C to ensure full conversion. The mixture was allowed to reach rt and was stirred for another 4 h. GC-MS analysis showed no further conversion, therefore diiodomethane (2 mL, 24.8 mmol, 2.4 eq) was added at rt. The reaction mixture was quenched by addition of saturated aq. NH4Cl solution after stirring at rt for additional 38 h. The phases were separated and the aqueous layer was extracted three times with CH2Cl2. The combined organic layers were dried over MgSO4 and all volatiles were removed under reduced pressure. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 10:1 to 4:1) to yield a mixture of 41 and 42 (1.00 g, 88% 41/42: 8:1) as a yellow oil. Rf=0.36 (pentane/diethyl ether, 1:1) Major component: (41): 1H NMR (400 MHz, CDCl3) -5.46 (m, 1H), 5.08-5.02 (ddt, J=15.2, 8.3, 1.6 Hz, 1H), 3.51-3.38 (m, 2H), 1.65-1.63 (dd, J=6.5, 1.6 Hz, 3H), 1.10-1.05 (m, 1H), 0.75-0.67 (m, 1H), 0.62-0.55 (m, 1H), 0.53-0.42 (m, 1H), 0.34-0.26 (m, 1H), 0.22-0.17 (m, 1H), 0.11-0.05 (m, 1H). 13C NMR (100 MHz, CDCl3

(CH), 66.6 (CH2), 22.5 (CH), 19.4 (CH3), 11.2 (CH2). HRMS (ESI+): m/z [M+H]+ calcd for C7H13O: 113.0966; found: 113.0961.

(E)-(2-(prop-1-en-1yl)cyclopropyl)methyl pivalate (43)

Pivalic acid (1.37 g, 13.4 mmol, 1.5 eq) was dissolved in 15 mL CH2Cl2 and cooled to 0 °C. To this solution were added DCC (5.55 g, 26.9 mmol, 3.0 eq) and DMAP (109 mg, 0.89

mmol, 0.1 eq) and the reaction mixture was stirred for 5 min at 0 °C before a solution of alcohols 41 and 42 (1.00 g, 8.96 mmol, 1.0 eq) in 5 mL CH2Cl2 was added. The reaction mixture was allowed to reach rt and was stirred at that temperature for 24 h. The reaction was worked up by filtration over celite and washing with pentane. The crude product was purified by column chromatography (SiO2, pentane/MTBE 10:1 to 2:1) to yield ester 43 as a yellow oil. (1.03 g, 5.25 mmol, 59%). Rf=0.77 (pentane/MTBE, 1:1) 1H NMR (500 MHz, CDCl3 -5.43 (m, 1H), 5.04-5.02 (ddt, J=15.1, 8.1 Hz, 1H), 3.93-3.84 (m, 2H), 1.58 (d, J=6.4 Hz, 3H), 1.15 (s, 9H), 1.05 (m,

OH OH+

O

O

Synthesis of a phosphoglycolipids based on lactobacillic acid

61

1H), 0.72 (m, 1H), 0.61-0.57 (m, 1H), 0.55-0.52 (m, 1H), 0.42 (m, 1H), 0.23 (m, 1H), 0.13 (m, 1H), 0.02 (m, 1H). 13C NMR (100 MHz, CDCl3

(CH), 67.6 (CH2), 34.2 (C), 27.1 (3 x CH3), 19.4 (CH3), 18.7 (CH), 17.6 (CH), 11.4 (CH2). HRMS was not measured.

(E)-(2-(prop-1-en-1yl)cyclopropyl)methyl pivalate (44)28

Olefin 43 (100 mg, 0.51 mmol, 1.0 eq) was dissolved in 10 mL of EtOH at rt and stirred vigorously. An atmosphere of oxygen was applied (balloon, 1 atm). Then, via syringe

pump, simultaneously a solution of riboflavin catalyst 43 (0.05 mmol, 0.1 eq) in EtOH (5 mL) and hydrazine hydrate (5.0 mmol, 10 eq) were added slowly over a period of 10 h. The reaction mixture was stirred at rt for another 6 h and the progress of the reaction was followed by TLC. CH2Cl2 was added, the mixture was washed three times with water, the aqueous layers were extracted with CH2Cl2, and the organic layers combined. After drying with Na2SO4, volatiles were removed in vacuo to yield the crude product. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 50:1 to 9:1) to yield ester 44 as a yellow oil (85 mg, 0.43 mmol, 84%). Rf=0.69 (pentane/diethyl ether, 4:1) 1H NMR (500.0 MHz, CDCl3

ppm 8.01 (s, 1H), 5.55-5.46 (m, 1H), 5.08-5.02 (ddt, J=15.2, 8.3, 1.6, 1H), 3.51-3.38 (m, 2H), 1.65-1.63 (dd, J=6.5, 1.6, 3H), 1.10-1.05 (m, 1H), 0.75-0.67 (m, 1H), 0.62-0.55 (m, 1H), 0.53-0.42 (m, 1H), 0.34-0.26 (m, 1H), 0.22-0.17 (m, 1H), 0.11-0.05 (m, 1H). 13C NMR (100.6 MHz, CDCl3 2), 38.7 (C), 27.2 (CH2), 27.1 (3 x CH3), 22.5 (CH2), 17.2 (CH), 13.9 (CH3), 10.0 (CH2). HRMS (ESI+): m/z [M+Na]+ calcd for C7H12O: 221.1517; found: 221.1512.

O

O

Chapter 3

62

3.6.3 A model substrate for the desired gycerolphospholipid21,23,29

(R)-3-((tert-butyldimethylsilyl)oxy)-2-(stearoyloxy)propyl oleate (48)

Oleic acid (750 mg, 2.65 mmol, 1 eq) was dissolved in 2 mL diethyl ether and Co[salen]-

catalyst 46 (16 mg, 0.026 mmol, 1 mol%) was added. This mixture was stirred in an open flask for 15 min at room temperature. Then the reaction mixture was flushed with nitrogen and DIPEA (0.46 mL, 2.78 mmol, 1 eq) and protected glycidol 26 (0.56 mL, 2.65 mmol, 1 eq) were added. The reaction mixture was stirred at rt overnight till the reaction was complete (followed by TLC eluent n-pentane/Et2O 1:1). As a work-up, DIPEA and diethyl ether were completely removed from the reaction mixture by evaporation under vacuum. Crude product 47 was dissolved in 10 mL n-heptane and to this was added DCC (656 mg, 3.18 mmol, 1.2 eq), DMAP (32 mg, 0.26 mmol, 10 mol%) and stearic acid (905 mg, 3.18 mmol, 1.2 eq). The reaction mixture was stirred at rt for 15 h. For purification the reaction mixture was directly put on the column. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 9:1 to 3:1) to yield the protected lipid 48 as a colorless oil (1.35 g, 1.84 mmol, 69%). 1H NMR (500.0 MHz, CDCl3 5.40–5.29 (m, 2H), 5.11-5.06 (m, 1H), 4.35 (dd, J=11.8, 3.7 Hz, 1H), 4.21–3.98 (m, 2H), 3.72 (d, J=6.2 Hz, 1H), 2.33–2.29 (m, 4H), 2.04– 1.99 (m, 4H), 1.68-1.54 (m, 4H), 1.24 (br s, 24H), 0.88 (t, 6H), 0.10 (s, 6H), 0.06 (s, 9H). 13C NMR (100.6 MHz, CDCl3 173.4 (C), 173.3 (C), 130.0 (CH), 129.7 (CH), 71.6 (CH), 65.3 (CH2), 62.4 (CH2), 34.3(CH2), 34.2 (CH2), 34.1 (CH2), 31.9 (2xCH2), 29.7 (CH2), 29.6 (4xCH2), 29.5 (CH2), 29.4 (2xCH2), 29.3 (3xCH2), 29.2 (3xCH2), 29.1 (2xCH2), 27.2 (2xCH2), 25.7 (3xCH3), 24.9 (2xCH2), 22.7 (CH2), 18.2 (CH2), 18.0 (CH2), 14.1 (CH3), -4.8 (CH3), -5.5 (CH3). HRMS was not measured.

(S)-3-hydroxy-2-(stearoyloxy)propyl oleate (49)

The protected lipid 48 (630 mg, 0.85 mmol, 1 eq) was solved in 5 mL acetonitrile under nitrogen and BF3·Et2O (243

mg, 1.71 mmol, 2 eq) was added. The reaction mixture was stirred at rt for 3 h (progress followed by TLC), and then water and CH2Cl2 were added. The layers were separated and the water layer was extracted three times with CH2Cl2. The combined organic layers where dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The deprotected lipid 49 was isolated as crude product (486

OTBDMS

O

O

O

O

OH

O

O

O

O

Synthesis of a phosphoglycolipids based on lactobacillic acid

63

mg, 0.78 mmol, 91%). 1H NMR (400.0 MHz, CDCl3 5.37–5.29 (m, 2H), 5.11-5.06 (m, 1H), 4.35-4.11 (m, 3H), 3.72 (d, J=6.2 Hz, 1H), 2.36–2.29 (m, 4H), 2.03– 1.98 (m, 4H), 1.64-1.59 (m, 4H), 1.24 (br s, 24H), 0.88 (t, 6H). 13C NMR (100.6 MHz, CDCl3

ppm 173.7 (C), 173.4 (C), 130.0 (CH), 129.7 (CH), 72.1 (CH), 62.0 (CH2), 61.5 (CH2), 34.3 (CH2), 34.1 (CH2), 31.9 (CH2), 31.9(CH2), 29.7 (CH2), 29.6 (4xCH2), 29.5 (CH2), 29.4 (2xCH2), 29.3 (3xCH2), 29.2 (3xCH2), 29.1 (2xCH2), 27.2 (2xCH2), 24.9 (2xCH2), 22.7 (CH2), 18.2 (CH2), 18.0 (CH2), 14.1 (CH3). HRMS was not measured.

2-((hydroxy((R)-3-(oleoyloxy)-2-(stearoyloxy)propoxy)phosphoryl)oxy)-N,N,N-trimethylethanaminium (50)

The deprotected lipid 49 (486 mg, 0.78 mmol, 1 eq) was dissolved in 10 mL dry THF

and cooled to 0 °C with an ice bath. To this were added triethylamine (0.33 mL, 2.34 mmol, 3 eq) and dioxaphospholane 2-oxide (333 mg, 2.34 mmol, 3 eq). Then the ice bath was removed and the reaction mixture was stirred at rt for 3 h (progress followed by TLC). The reaction mixture was filtered over celite and the filter cake was washed with CH2Cl2. After evaporation of the solvent the crude product was directly dissolved in 5 mL acetonitrile in a pressure tube under nitrogen. The tube was cooled with an acetone/dry ice bath to -78 °C. In the pressure tube 2 mL of trimethylamine was condensed. After the addition the pressure tube was closed as fast as possible. The reaction mixture was allowed to reach rt, and then the reaction was stirred at 70 °C for 4 h (progress followed by TLC). The reaction mixture was cooled to rt and poured into a petri dish. The excess of trimethylamine was allowed to evaporate. The crude product was purified by column chromatography (SiO2, chloroform/methanol 3:1) to yield phosphorcholine lipid 50 as white solid (87.7 mg, 0.11 mmol, 14%). HRMS (ESI+): m/z [M]+ calcd for C44H87NO8P+: 788.6164; found: 788.6148.

O

O

O

O

O

PO

ON

O

Chapter 3

64

3.7 References

1. B. Alberts, A. Johnson, J. Lewis, M. Raff, R. Keith, P. Walter, Molecular Biology of the Cell, 5th edition, Taylor & Francis, 2007.

2. N. Levina, S. Tötemeyer, N. R. Stokes, P. Louis, M. A. Jones, and I. R. Booth, EMBO J., 1999, 18, 1730–1737.

3. C. C. Häse, A. C. L. Dain, and B. Martinac, J. Biol. Chem., 1995, 270, 18329–18334.

4. G. Chang, R. H. Spencer, A. T. Lee, M. T. Barclay, and D. C. Rees, Science, 1998, 282, 2220–2226.

5. J. P. Birkner, B. Poolman, and A. Koçer, Proc. Natl. Acad. Sci., 2012, 109, 12944–12949.

6. K. Hofmann and R. A. Lucas, J. Am. Chem. Soc., 1950, 72, 4328–4329.

7. G. I. Veld, A. J. M. Driessen, J. A. F. Op den Kamp, and W. N. Konings, Biochim. Biophys. Acta, 1991, 1065, 203–212.

8. K. Hofmann, O. Jucker, W. R. Miller, A. C. Young, and F. Tausig, J. Am. Chem. Soc., 1954, 76, 1799–1804.

9. K. Hofmann, G. J. Marco, and G. A. Jeffrey, J. Am. Chem. Soc., 1958, 80, 5717–5721.

10. G. D. Coxon, J. R. Al-Dulayymi, M. S. Baird, S. Knobl, E. Roberts, and D. E. Minnikin, Tetrahedron Asymmetry, 2003, 14, 1211–1222.

11. S. Kobayashi, R. Tokunoha, M. Shibasaki, R. Shinagawa, and K. Murakami-Murofushi, Tetrahedron Lett., 1993, 34, 4047–4050.

12. H. Suematsu, S. Kanchiku, T. Uchida, and T. Katsuki, J. Am. Chem. Soc., 2008, 130, 10327–10337.

13. K. Geurts, S. P. Fletcher, and B. L. Feringa, J. Am. Chem. Soc., 2006, 128, 15572–15573.

14. S. R. Harutyunyan, T. den Hartog, K. Geurts, A. J. Minnaard, and B. L. Feringa, Chem. Rev., 2008, 108, 2824–2852.

15. A. Fürstner, K. Radkowski, C. Wirtz, R. Goddard, C. W. Lehmann, and R. Mynott, J. Am. Chem. Soc., 2002, 124, 7061–7069.

16. C. W. Lee and R. H. Grubbs, Org. Lett., 2000, 2, 2145–2147.

17. A. Fürstner, O. R. Thiel, and L. Ackermann, Org. Lett., 2001, 3, 449–451.

Synthesis of a phosphoglycolipids based on lactobacillic acid

65

18. C. Smit, M. W. Fraaije, and A. J. Minnaard, J. Org. Chem., 2008, 73, 9482–9485.

19. J. F. Teichert, T. den Hartog, M. Hanstein, C. Smit, B. ter Horst, V. Hernandez-Olmos, B. L. Feringa, and A. J. Minnaard, ACS Catal., 2011, 1, 309–315.

20. E. N. Jacobsen, Acc. Chem. Res., 2000, 33, 421–431.

21. L. P. C. Nielsen, C. P. Stevenson, D. G. Blackmond, and E. N. Jacobsen, J. Am. Chem. Soc., 2004, 126, 1360–1362.

22. P. Fodran and A. J. Minnaard, Org. Biomol. Chem., 2013, 11, 6919–6928.

23. V. Hornillos, F. Amat-Guerri, and A. U. Acuña, J. Photochem. Photobiol. Chem., 2012, 243, 56–60.

24. E. Casas-Arce, B. ter Horst, B. L. Feringa, and A. J. Minnaard, Chem. Eur. J., 2008, 14, 4157–4159.

25. B. Chen, Z. Lu, G. Chai, C. Fu, and S. Ma, J. Org. Chem., 2008, 73, 9486–9489.

26. L. J. Stuart, J. P. Buck, A. E. Tremblay, and P. H. Buist, Org. Lett., 2006, 8, 79–81.

27. C. R. Theberge, C. A. Verbicky, and C. K. Zercher, J. Org. Chem., 1996, 61, 8792–8798.

28. C. Smit, M. W. Fraaije, and A. J. Minnaard, J. Org. Chem., 2008, 73, 9482–9485.

29. B. ter Horst, C. Seshadri, L. Sweet, D. C. Young, B. L. Feringa, D. B. Moody, and A. J. Minnaard, J. Lipid Res., 2010, 51, 1017–1022.

Chapter 3

66

Copper(I)-catalyzed asymmetric alkylation of aldehydes with Grignard reagents: A direct access to secondary alcohols

In this chapter the asymmetric copper-catalyzed 1,2-addition of Grignard reagents to aryl aldehydes and , -unsaturated aldehydes is described. Due to a higher reactivity of aldehydes compared to ketones, a lower enantioselectivity was observed, therefore different additives have been studied to enhance the enantioselectivity of the addition reaction.

Chapter 4

68

4.1 Introduction

The asymmetric 1,2-addition of organometallic reagents to aldehydes is a fundamental reaction to form enantioenriched secondary alcohols. The great interest to synthesize enantioenriched secondary alcohols derives from occurrence of this moiety in many natural products, fragrances and biological active compounds.1 The first attempts to perform this alkylation with Grignard reagents in an asymmetric manner were carried out in the presence of stoichiometric amounts of chiral auxiliaries or chiral modifiers.2–4 In 1984, Oguni et al. reported the first catalytic enantioselective alkylation of benzaldehyde with diethylzinc as organometallic reagent and different amino alcohols as chiral catalyst (Scheme 4.1).5

O 2-amino-1-alcohol, (Et)2Zntoluene, 20 °C, 48 h OH

yield ee

OHNH2

(S)-leucinol

OHNH2

(S)-valinol

OHNH2

(S)-phenylalaninol

OHNH2

(S)-alaninol

HN

HO

(S)-prolinol

96%49%

95%47%

98%40%

97%26%

100%28%

*H

Scheme 4.1. First catalytic enantioselective addition of an organometallic reagent to benzaldehyde.

Further developments in the enantioselective addition of dialkylzinc reagents to aldehydes were reported by Noyori and co-workers. Their studies focused on the mechanism and showed that less reactive organometallic reagents than magnesium or lithium reagents were essential to reach high enantioselectivities in the alkylation of aldehydes (Scheme 4.2).6

O 2 mol% DAIB, (Et)2Zntoluene, 0 °C, 6-24 h OH

yield 80-97%ee 61-98%

R R

OHNMe2

DAIB

*H

Scheme 4.2. Developments towards a catalytic system for the alkylation of aldehydes.

A drawback in the use of dialkylzinc reagents is the thermal instability of their higher homologs. One option to prepare diorganozinc reagents in situ is by transmetallation of Grignard reagents to zinc salts (ZnX2, X: Cl, Br and OMe). This method was

A direct access to secondary alcohols

69

applied by Seebach et al. in the addition of dialkylzinc compounds to aldehydes in the presence of a spirotitanium complex as chiral catalyst (Scheme 4.3).7

R1 CH2X2

1) 2 eq Mg in Et2O2) ZnCl23) 2 eq 1,4-dioxane (separation of precpitate)4) 0.15 eq spirotitanate 5) 1.2 eq Ti(OCHMe2)4 78 °C, 1 h6) 1.0 eq R2CHO 78 °C to 30 °C, 15-24 h7) work-up

R2 R1OH

OH OH OH OH

yieldee

40%98%

>98%70%

40%90%

69%94%

O

O OTi

OPh Ph

Ph Ph

spirotitanate

O

O

O

OPhPh

Ph Ph

Scheme 4.3. Application of in situ formed dialkylzinc reagents in the alkylation of aldehydes.

In this one-pot procedure it is crucial to separate the precipitated magnesium salts before using the solution in the alkylation, otherwise a low enantioselectivity is observed due to the competing racemic reaction promoted by these salts. As a further optimization of this procedure, Seebach and co-workers reported in 1994 the direct use of Grignard reagents in the transmetallation to Ti(OiPr)3Cl forming the less reactive organotitanium reagent, which then was used in combination with Ti-TADDOLate as catalyst for the asymmetric 1,2-addition to aldehydes.8 A significant improvement was described by Harada et al. since in their catalytic system the magnesium salts from the transmetallation could remain in the R-Ti(OiPr)3 solution and still high enantioselectivities were reached (Scheme 4.4).9,10

Chapter 4

70

R1 MgX Ti(OiPr)4+R2 H

O

R2 R1

OH+ OH

OH

Ph

Ph

2 mol% DPP-BINOL1.4 eq Ti(OiPr)4

CH2Cl2, 0 °C, 3 h

(2.2 eq) (4.4 eq)

DPP-BINOL

OH

nBu

OH

nBu

OH

nBu

F3C

HO nBu

yieldee

86%93%

82%95%

62%95%

90%96%

Scheme 4.4. Optimized procedure for the asymmetric alkylation of aldehydes by Harada et al..

Additional studies by Da et al. showed that the amount of Ti(OiPr)4 can be reduced to 0.9 eq by using a chelating additive bis[2-(N,N’-dimethylamino)ethyl] ether (BDMAEE), that makes the removal of magnesium salts unnecessary.11,12 With this catalytic system a variety of Grignard reagents were tolerated and the scope of the aromatic aldehydes could be broadened (Scheme 4.5).

R1 MgX+R2 H

O

R2 R1

OH

15 mol% BINOL0.9 eq Ti(OiPr)42.0 eq BDMAEEMTBE, 0 °C, 5 h

(2.0 eq)

yieldee

46-91%35->99%

OHOH

(S)-BINOL

NO

N

BDMAEE

Scheme 4.5. Enantioselective alkylation of aldehydes with BDMAEE as chelating agent.

In the latest study of Yus et al. a new catalytic system based on Ti(OiPr)4, an Ar-BINMOL ligand (L1) and various Grignard reagents, in particular MeMgBr, was reported. This catalytic system allows not only alkylation to aromatic aldehydes, also aliphatic aldehydes undergo the reaction with high yields and high enantioselectivities (Scheme 4.6).13,14

A direct access to secondary alcohols

71

MeMgX+R2 H

O

R2 Me

OH

20 mol% L110 eq Ti(OiPr)4

Et2O, 20 °C, 3 h

(2.5 eq)

Ar-BINMOL(Sa,R)-L1

OH

N

HO H

OH OH OH OH

yieldee

61%92%

99%83%

58%99%

98%88%

Scheme 4.6. Latest development in the alkylation of aliphatic aldehydes with MeMgBr.

In the asymmetric alkylation with titanium organyls and catalytic amounts of ligands, high enantioselectivities are reached, because of well-defined intermediates in which the alkyl group can be transferred specifically to one face of the carbonyl double bond. Still a major drawback is the huge access of Ti(OiPr)4 used in most of the protocols. Thus, we saw the need to develop a new catalytic system, in which only catalytic amounts of ligand and metal precursors are used. Recently we reported a new procedure for the asymmetric copper(I)-catalyzed 1,2-addition of Grignard reagents to -unsaturated ketones.15,16 Those optimized conditions were applied in

-unsaturated aldehydes (Scheme 4.7).17

OH

Me

O

Me

5 mol% CuBr.SMe26 mol% L2

1.3 eq iBuMgBrMTBE, 78 °C, 12 h

H

1

FePP

H3C H

rev-JosiPhos (S, RFe)-L2

2yield 92%ee 60%

Scheme 4.7. Application of the copper- -unsaturated aldehydes.

4.2 Goal of this study

In this study we applied the optimized reaction conditions for the 1,2-addition of Grignard reagents to aryl ketones, in the corresponding reaction to aryl aldehydes

-unsaturated aldehydes. In the preliminary studies (Scheme 4.7), an

Chapter 4

72

enhancement of the ee was observed when 15 mol% iPrOH was used as an additive in the -unsaturated aldehydes.17 We tried to understand this observation and tested, whether the same effect was observed for aryl aldehydes.

4.3 Results and discussion

In the first set of experiments (Table 4.1, entries 1 and 2), the conditions, as previously described, were used in the 1,2-addition to aldehyde 1. The observed yields and enantioselectivities are in the same range as previously described (Scheme 4.7).17

Table 4.1. Repeating conditions of preliminary study and further optimization with aldehyde 1.

OH

Me

O

Me

5 mol% CuBr.SMe26 mol% L2

1.3 eq iBuMgBrMTBE, T, 12 h

H

1

FePP

H3C H

rev-JosiPhos (S, RFe)-L2

2

entry product T solvent additive eea ratio [%]b

1:2

1 2 78 °C MTBE no 59% 9:91

2 2 78 °C MTBE 20 mol% iPrOHc 81% 14:86

3 2 78 °C diisopropyl ether

no 55% 1:99

4 2 78 °C CH2Cl2 no 29% 4:96

5 2 83 °C MTBE no 51% 6:94

a The enantiomeric excess was determined by HPLC, Chiralcel OD-H column, n-heptane/i-PrOH 99:1, 40 °C isotherm, detection at 254 nm, retention time (min): 17.9 (major) and 20.0 (minor). b Ratio determined by GC-MS analysis. c 1.8 eq Grignard reagent.

The influence of the solvent was studied, to increase the enantioselectivity (entries 3 and 4). The reaction in diisopropyl ether gave similar results and the reaction with dichloromethane as solvent gave only 29% ee, therefore MTBE was used as solvent

A direct access to secondary alcohols

73

for the further screening. It was tried to increase the ee by lowering the temperature to -83 °C (entry 5), but this had no significant influence. In all reactions in this study, generally around 5% of the starting material was not converted. This might be due to the formation of a stable copper(III)-complex, which then does not further transfer the alkyl group. The nature of the copper(III) intermediates are studied by rapid injection NMR in the group of Ogle et al..18

In the subsequent reactions, the optimized conditions were applied to aryl aldehyde 3. In this case, around 20% of 1,2-reduction product 5 was observed as side product (see ratio of starting material to products 4 and 5, Table 4.2). This side product is formed due to -hydride transfer from the Grignard reagent. This problem occurs especially with branched Grignard reagents.

Table 4.2. Optimizing the reaction conditions with aryl aldehyde 3.

Alk

OHO5 mol% CuBr.SMe2

6 mol% L21.3 eq Grignard

MTBE, 78 °C, 12 hH

3 4a: 2-ethylbutyl4b: iso-butyl

+

OH

5

entry product solvent additive eea ratio [%]b

3:4:5

1 4a OH

no 33% 35:50:15

2 4a OH

15 mol% iPrOHc 34% 88:9:3

3 4a OH

30 mol% iPrOHd 37% 5:75:20

4 4a

OH

60 mol% iPrOHe 37% 8:71:21

Chapter 4

74

entry product solvent additive eea ratio [%]b

3:4:5

5 4b

OH

no 30% 7:77:16

6 4b OH

20 mol% iPrOHf 12% n.d.

7 4b OH

10 mol% LiCl 27% n.d.

8g 4b OH

no 32% 4:79:17

9h 4b OH

no 27% n.d.

a The enantiomeric excess was determined by HPLC, Chiralpak AD-H column, n-heptane/i-PrOH 98:2, 40 °C isotherm, detection at 210 nm, retention time (min): 18.6 (major) and 21.1 (minor). b Ratio determined by GC-MS analysis. c 1.8 eq Grignard reagent. d 2.3 eq Grignard reagent. e 3.2 eq Grignard reagent. f 2.0 eq Grignard reagent. g Reverse addition, addition of aldehyde to the reaction mixture. h Diluted reaction mixture c = 0.018 M; standard concentration: c = 0.075 M

In the first set of reactions, 2-ethylbutylmagnesium bromide was used in the addition to aldehyde 3 (entries 1-4). The ee of alcohol 4a is 30% lower compared to alcohol 2, and in addition there was no enhancement of the ee observed when iPrOH was added to the reaction mixture. Even the amount of additive had no influence on the enantioselectivity. Especially, the reaction in entry 2 shows only 9% conversion. To exclude that the decrease in ee resulted from the Grignard reagent used, the reactions were repeated with iso-butylmagnesium bromide to give the secondary alcohol 4b (entries 5 and 6). The ee in the standard reaction stays in the range of 30%, but in the reaction with iPrOH as additive only 12% ee was observed. Further attempts to increase the enantioselectivity, by breaking the magnesium aggregates using LiCl as an additive, failed. The inverse addition, that is, the addition of aldehyde 3 to the reaction mixture containing the full amount of Grignard reagent, gave only 32% ee. By changing the concentration of the reaction mixture to a more dilute solution a similar ee of 27% was observed.

A direct access to secondary alcohols

75

In the screening of various aryl aldehydes, aldehyde 6 gave the highest ee with 51%, therefore the reaction conditions were optimized using this substrate. First, diisopropyl ether and dichloromethane were used as solvents, but in both cases no increase in enantioselectivity was observed.

Table 4.3. Further investigations for optimal reaction conditions using aldehyde 6.

O HO5 mol% CuBr.SMe2

6 mol% L1.3 eq iBuMgBr

MTBE, 78 °C, 12 h

FePP

H3C H

L4

OO

P N

L5L3

FeP

PCH3

H

HO

+

6 7 8

H

entry product ligand catalyst loading additive eea ratio [%]b

6:7:8

1 7 L2 6 mol% no 51% 11:80:9

2 7 L2 6 mol% DME

20 mol% 33% 10:76:14

3 7 L2 6 mol% BDMAEE 20 mol%

43% 11:76:13

4 7 L2 20 mol% no 76% 8:87:5

5c 7 L2 6 mol% no 51% 4:79:17

6 7 L3 6 mol% no 19% n.d.

7 7 L4 6 mol% no 21% n.d.

8 7 L5 6 mol% no rac n.d.

a The enantiomeric excess was determined by HPLC, Chiralcel OJ-H column, n-heptane/i-PrOH 98:2, 40 °C isotherm, detection at 215 nm, retention time (min): 20.8 (minor) and 25.4 (major). b Ratio determined by GC-MS analysis. c Reverse addition, addition of aldehyde to the reaction mixture.

Chapter 4

76

Furthermore 1,2-dimethoxyethane (DME) and BDMAEE were used as additives to chelate free magnesium salts to prevent the racemic background reaction (Table 4.3, entries 2 and 3). This turned out to be not the case; enantioselectvities between 30 to 40% were reached. An increase of ee was observed only when 20 mol% of the catalyst was used (entry 4). As previously for aldehyde 3, also for aldehyde 6 a reverse addition gave no increase of enantioselectivity. Additionally, two other ferrocene ligands, L3 and L4, and one phosphoramidite ligand, L5, were investigated (entries 6-8). L2 gave the highest ee in this set of ligands. For this reason the further screening of the reaction was performed with L2 in MTBE at - 8 °C.

In the subsequent set of experiments, different aryl aldehydes were applied, which either have a different structure or carry an electron-withdrawing or -donating substituent on the phenyl ring. In addition we tested also 2-thiophenecarboxaldehyde as an example of a heterocyclic compound (Table 4.4). In the case of the unsubstituted aromatic aldehyde 9a and the two naphthyl aldehydes 9b, 9c (Table 4.4, entries 1-3) one can see an increase in ee, with increasing bulk close to the carbonyl double bond. In contrast to this is the low ee of compound 10d (entry 4), where one face should be more shielded by the methyl group in ortho position. The highest ee is observed with substrate 9g (entry 7), the low yield of this product is due to the volatility of the compound.

Table 4.4. Screening different aryl aldehydes

OHO5 mol% CuBr.SMe2

6 mol% L21.3 eq Grignard

MTBE, 78 °C, 14 hH

9 10

+

OH

11

R R R

entry Substrate yielda eeb

1 9a O

H

10a 30% 27%

2 9b O

H

10b 18% 45%

3 9c

O H

10c 44% 52%

A direct access to secondary alcohols

77

entry Substrate yielda eeb

4 9d O

H

10d 40% 26%

5 9e O

F

H

10e 48% 43%

6 9f

O

CF3

H

10f 30% 23%

7 9g SO

H

10g 6% 67%

a Isolated yield after work-up and purification by column chromatography. b The enantiomeric excess was determined by HPLC, for details see experimental section.

The enantioselectivities of the secondary alcohols derived from aryl aldehydes are between 23-67% due to the fast blank reaction. For the blank reaction substrate 9d was used, and after 14 h 75% of conversion to alcohol 10d was observed. Thus, the focus was turned again to -unsaturated aldehydes since in the initial studies high enantioselectivities were obtained with these substrates (Scheme 4.7). For an easier

-unsaturated aldehydes 15g and 15h were synthesized in three steps (Scheme 4.8). In the first step aldehyde 9 was reacted with the commercially available HWE reagent 12 -unsaturated ester 13, for both esters 13g and 13h only the E-double bond isomer was identified by NMR. The ester was reduced with DIBAlH to alcohol 14, which was subsequently

-unsaturated aldehyde 15.

Chapter 4

78

R

O O

OR

DIBAlH, Et2O-83°C to -40 °C, 2 d

NMO, TPAP DCM, rt, 14 h

OH

R

O

S

PO

O

O n-BuLi, THF-10°C to rt, on

+

14

15g24%

EtOEtO

O

12 13

15h28%

58%

41%over two steps

overall yield

9g9h

R: thiophene R: m-toluol

H

O

R

15

H

Scheme 4.8. -unsaturated aldehydes.

Next, the synthesized -unsaturated aldehydes were tested in the asymmetric 1,2-addition of Grignard reagents applying two different sets of conditions (Scheme 4.9). With condition 1, standard conditions are meant and in condition 2, 20 mol% iPrOH was used as an additive and the Grignard reagent was added slowly.

OH

S

OH

condition 1: 88% conversion, 64% eecondition 2: 86% conversion, 62% ee

RR

5 mol% CuBr.SMe26 mol% L2

1.3 eq iBuMgBrMTBE, 78 °C, 14 h

15g15h

16g16h

OH

H

O

condition 1: 89% conversion, 40% ee

16h 16g

R: thiophene R: m-toluol

R: thiophene R: m-toluol

Scheme 4.9. Copper(I)-catalyzed 1,2- -unsaturated aldehydes.

The results shown in Scheme 4.9 are preliminary results. The conversion was determined by GC-MS analysis, and the compounds were not further characterized by NMR or HRMS. The enantiomeric excess was determined by HPLC. For aldehyde 15h, in both cases a similar ee of 64% was reached, which means also for this substituted -unsaturated aldehyde no enhancement of ee was observed when

A direct access to secondary alcohols

79

iPrOH was added to the reaction mixture. Surprisingly, substrate 15g gave even a lower ee than its matching aryl aldehyde.

4.4 Conclusion

We investigated the copper(I)-catalyzed 1,2-addition of Grignard reagents to aryl aldehydes. In comparison to the aryl ketones, for the aryl aldehydes lower enantioselectivities were observed (30-60% ee). The selectivity decreases due to the higher reactivity of aldehydes in comparison to ketones. In addition, the magnesium salts (MgBr2) in the reaction mixture catalyze the blank reaction, this phenomenon was first described by Harada et al..10 Our attempts to chelate the free magnesium salts or to break the magnesium clusters with LiCl showed also no effect on the ee. No influence on the enantioselectivity was observed, when the aldehyde was added slowly to the reaction mixture. This supports the effect of a fast blank reaction, since in the reaction one drop of the aldehyde solution should be surrounded by more molecules of the chiral catalyst than in the addition the other way around.

The enhancement of ee when iPrOH was used, was only observed for aldehyde 1. The enantioselectivity seems to be independent of the amount of iPrOH used in the reaction, as we showed for aldehyde 3. The conversion of the substrates was never complete, there was always around 4-10% starting material left, this might be due to the formation of a stabilized copper(III) intermediate.18 In reactions of aryl aldehydes all the time there was 1,2-reduction product as by-product observed.

In summary the best set of conditions for the 1,2-addition of aldehydes with Grignard reagents is the copper-complex with L2 in MTBE at - 8 °C. Future studies should investigate the mechanism (Figure 4.1), to understand the coordination of copper to the carbonyl double bond. One option to study the activated complex or copper intermediates, would be rapid injection NMR. With this technique, Ogle et al. showed a difference in coordination depending on the copper source, and the coordination of copper to the double bond in a methyl vinyl ketone.18,19 Based on those investigations new ligands can be designed, to reach a higher differentiation of the two faces of the carbonyl double bond to increase the enantioselectivity. Another crucial point is the prevention of the activation of the carbonyl double bond by the free magnesium salts.

OH

PhMg

Br

Br

CuR2P PR2

R

Figure 4.1. Proposed transition state of the reaction.

Chapter 4

80

4.5 Experimental

For general information see experimental of chapter 2.

4.5.1 Alkylation of aryl aldehydes

General procedure for the copper-catalyzed 1,2-addition

2 (0.015 mmol, 5 mol%) and (S,RFe)-reverse Josiphos (L2) (0.018 mmol, 6 mol%) were dissolved in dry MTBE (4 mL) and stirred at room temperature for 15 min. To this mixture was added the corresponding aldehyde (0.6 mmol, 1 eq) in 4 mL dry MTBE. Then the mixture was cooled to 78 °C. After stirring for 15 min at 78 °C, the corresponding Grignard reagent (0.72 mmol, 1.2 eq) was added over 30 min. The reaction mixture was stirred at 78 °C for 14 h. The reaction was quenched with 1 mL MeOH and 2 mL saturated aq. NH4Cl. After the reaction mixture reached rt, the layers were separated and the water layer was extracted three times with CH2Cl2. The combined organic layers were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. From the crude product a sample was taken for GC-MS analysis to determine the ratio of the 1,2-addition product, reduction product and starting material. The enantiomeric ratio was determined after purification by column chromatography.

3-methyl-1-phenylbutan-1-ol (10a)11

The title compound was prepared from aldehyde 9a following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1) afforded 10a as a light yellow oil (29.6 mg, 0.180 mmol, 30%, 27% ee). 1H NMR (400.0 MHz, CDCl3): ppm

7.35-7.19 (m, 5H), 4.20-4.10 (m, 1H), 1.75-1.57 (m, 3H), 1.46-1.40 (m, 1H), 0.90 (dd, 6H). 13C NMR (100.6 MHz, CDCl3): ppm 144.2 (C), 128.5 (3xCH), 125.8 (2xCH), 72.8 (CH), 48.3 (CH2), 24.8 (CH), 23.1 (CH3), 22.2 (CH3). The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OD-H column, n-heptane/i-PrOH 95:5, 40 °C isotherm, detection at 215 nm, retention time (min): 12.8 (minor) and 13.2 (major).

3-methyl-1-(naphthalen-1-yl)butan-1-ol (10b)11

The title compound was prepared from aldehyde 9b following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1) afforded 10b as yellow oil (23.4 mg, 0.109 mmol, 18%, 45% ee). 1H NMR (400.0 MHz,

OH

OH

A direct access to secondary alcohols

81

CDCl3): ppm 7.85-7.82 (m, 3H), 7.79 (s, 1H), 7.51-7.49 (m, 3H), 4.93 (m, 1H), 1.88 (s, 1H), 1.86-1.79 (m, 2H), 1.78-1.71 (m, 1H), 0.99-0.97 (dd, 6H). 13C NMR (100.6 MHz, CDCl3): ppm 142.5 (C), 133.3 (C), 132.9 (C), 128.3 (CH), 127.9 (CH), 127.7 (CH), 126.1 (CH), 125.8 (CH), 124.5 (CH), 124.1 (CH), 72.9 (CH), 48.2 (CH2), 24.9 (CH), 23.1 (CH3), 22.3 (CH3). The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OJ-H column, n-heptane/i-PrOH 98:2, 40 °C isotherm, detection at 214 nm, retention time (min): 45.8 (minor) and 52.5 (major).

3-methyl-1-(naphthalen-1-yl)butan-1-ol (10c)11

The title compound was prepared from aldehyde 9c following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1) afforded 10c as a light yellow oil (56.3 mg, 0.263 mmol, 44%, 52% ee). 1H NMR (400.0 MHz, CDCl3): ppm

8.11 (d, 1H), 7.88 (d, 1H), 7.78 (d, 1H), 7.66 (d, 1H), 7.55-7.46 (m, 3H), 5.57 (d, 1H), 2.00-1.92 (m, 2H), 1.85 (m, 1H), 1.72 (m, 1H), 0.99-0.97 (dd, 6H). 13C NMR (100.6 MHz, CDCl3): ppm 141.0 (C), 133.8 (C), 130.3 (C), 128.9 (CH), 127.8 (CH), 126.0 (CH), 125.5 (2xCH), 123.0 (CH), 122.0 (CH), 69.3 (CH), 47.7 (CH2), 25.3 (CH), 23.6 (CH3), 21.9 (CH3). The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OJ-H column, n-heptane/i-PrOH 98:2, 40 °C isotherm, detection at 254 nm, retention time (min): 20.1 (minor) and 24.1 (major).

1-(2,5-dimethylphenyl)-3-methylbutan-1-ol (10d)

The title compound was prepared from aldehyde 9d following the general. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1) afforded 10d as light yellow oil (45.8 mg, 0.238 mmol, 40%, 26% ee). 1H NMR (400.0 MHz, CDCl3):

ppm 7.30 (s, 1H), 7.00 (q, 2H), 4.99 (dd, 1H), 2.33 (s, 3H), 2.29 (s, 3H), 1.94-1.80 (m, 1H), 1.71-1.64 (m, 3H), 0.99 (dd, 6H). 13C NMR (100.6 MHz, CDCl3): ppm 143.3 (C), 135.8 (C), 130.9 (C), 130.3 (CH), 127.7 (CH), 125.6 (CH), 68.8 (CH), 47.6 (CH2), 25.0 (CH), 23.6 (CH3), 21.1 (CH3), 21.1 (CH3), 18.5 (CH3). HRMS was not measured. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralpak AD-H column, n-heptane/i-PrOH 98:2, 40 °C isotherm, detection at 210 nm, retention time (min): 18.8 (major) and 21.3 (minor).

HO

OH

Chapter 4

82

1-(4-fluorophenyl)-3-methylbutan-1-ol (10e)11

The title compound was prepared from aldehyde 9e following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1) afforded 10e as yellow oil (53.2 mg, 0.292 mmol, 48%, 43% ee). 1H NMR (400.0 MHz, CDCl3):

ppm 7.30 (dd, 2H), 7.02 (dd, 2H), 4.76-4.68 (m, 1H), 1.88 (br s, 1H), 1.76-1.61 (m, 2H), 1.53-1.40 (m, 1H), 0.94 (dd, 6H). 13C NMR (100.6 MHz, CDCl3): ppm 163.3 (C), 140.9 (C), 127.5 (CH), 127.4 (CH), 115.3 (CH), 115.1 (CH), 72.1 (CH), 48.4 (CH2), 24.7 (CH), 23.1 (CH3), 22.2 (CH3). The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel AS-H column, n-heptane/i-PrOH 100:0, 40 °C isotherm, detection at 254 nm, retention time (min): 44.3 (major) and 49.9 (minor).

3-methyl-1-(3-(trifluoromethyl)phenyl)butan-1-ol (10f)

The title compound was prepared from aldehyde 9f following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1) afforded 10f as light yellow oil (42.9 mg, 0.185 mmol, 30%, 23% ee). 1H NMR (400.0 MHz, CDCl3): ppm 7.61 (s, 1H), 7.52 (d, 2H), 7.48-7.41 (m, 1H), 4.82-4.79 (m, 1H), 2.06 (br s,

1H), 1.77-1.69 (m, 2H), 1.51-1.45 (m, 1H), 0.96 (dd, 6H). 13C NMR (100.6 MHz, CDCl3): ppm 146.3 (C), 130.6 (C), 129.2 (CH), 128.8 (CH), 122.6 (2xCH), 72.1 (CH), 48.4

(CH2), 24.7 (CH), 23.2 (CH3), 22.7 (CH3). HRMS was not measured. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OD-H column, n-heptane/i-PrOH 95:5, 40 °C isotherm, detection at 215 nm, retention time (min): 9.3 (minor) and 10.0 (major).

3-methyl-1-(thiophen-2-yl)butan-1-ol (10g)11

The title compound was prepared from aldehyde 9g following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1) afforded 10g as light yellow oil (6.5 mg,

0.038 mmol, 6%, 67% ee). 1H NMR (400.0 MHz, CDCl3): ppm 7.25 (d, 1H), 6.98-6.95 (m, 1H), 5.03-4.99 (m, 1H), 1.89 (br s, 1H), 1.84-1.61 (m, 3H), 0.96 (dd, 6H). 13C NMR (100.6 MHz, CDCl3): ppm 143.3 (C), 130.3 (CH), 127.7 (CH), 125.5 (CH), 68.8 (CH), 47.6 (CH2), 25.0 (CH), 23.6 (CH3), 21.7 (CH3). The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OD-H column, n-heptane/i-PrOH 95:5, 40 °C isotherm, detection at 233 nm, retention time (min): 12.7 (minor) and 13.6 (major).

OH

F

OH

CF3

SOH

A direct access to secondary alcohols

83

-unsaturated aldehydes (15)

General procedure for the synthesis of , -unsaturated ester (13)

O O

OP

O

O

O n-BuLi, THF-10°C to rt, on

+EtO

EtO

9 12 13R R

HWE reagent 12 (1.6 eq) was dissolved in 18 mL dry THF and cooled with an ice bath. At this temperature n-BuLi (1.6 M hexane solution, 1.3 eq) was added to the reaction mixture and stirred for 30 min. Then aldehyde 9 (1.0 eq) was added as solution in 2 mL THF. The reaction mixture was stirred overnight, while warming to rt. The reaction was quenched by adding saturated aq. NH4Cl solution. The layers were separated and the water layer was extracted three times with diethyl ether. The combined organic layers were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The -unsaturated ester 13 was isolated after purification by column chromatography.

(E)-ethyl 2-methyl-3-(thiophen-2-yl)acrylate (13g)

The title compound was prepared from aldehyde 9g (0.30 mL, 2.62 mmol) following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1)

afforded 13g as yellow oil (300 mg, 1.53 mmol, 58%). 1H NMR (400.0 MHz, CDCl3): ppm 7.75 (s, 1H), 7.35 (d, 1H), 7.16 (d, 1H), 6.99 (dd, 1H), 4.16 (q, 2H), 2.11 (s, 3H), 1.24 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 168.4 (C), 139.3 (C), 131.6 (CH), 131.4 (CH), 129.0 (CH), 127.3 (CH), 124.9 (C), 60.9 (CH2), 14.3 (CH3), 14.2 (CH3). HRMS was not measured.

(E)-ethyl 2-methyl-3-(m-tolyl)acrylate (13h)

The title compound was prepared from aldehyde 9h (0.47 mL, 4 mmol) following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1) afforded 13h as yellow oil (758 mg, 3.71 mmol, 93%). 1H NMR (400.0 MHz,

CDCl3): ppm 7.61 (s, 1H), 7.52 (d, 2H), 7.48-7.41 (m, 1H), 4.20 (q, 2H), 2.06 (s, 6H), 1.24 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 168.4 (C), 139.3 (C), 131.6 (CH), 131.4 (CH), 129.0 (CH), 127.3 (CH), 124.9 (C), 60.9 (CH2), 14.3 (CH3), 14.2 (CH3). HRMS was not measured.

O

O

O

OS

Chapter 4

84

General procedure for the synthesis of , -unsaturated aldehyde (15)

O

O

DIBAlH, Et2O78 °C, 16 h OH

RR

O

R

NMO, TPAP DCM, rt, 14 h

13 14 15 Ester 13 (1 eq) was dissolved in 5 mL dry diethyl ether and cooled to -50 °C. At this temperature DIBAlH (1 M THF solution, 2.2 eq) was added and the reaction mixture was stirred till all starting material was consumed (followed by TLC). The reaction was quenched by adding saturated aq. Rochelle’s salt solution (potsassium sodium tartrate) and this mixture was stirred for 1 h, for easier separation. The layers were separated and the water layer was extracted three times with diethyl ether. The combined organic layers were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. The crude -unsaturated alcohol 14 was immediately used in

-unsaturated aldehyde 15. Alcohol 14 (1 eq) was dissolved in dry DCM, and to this was added NMO (1.3 eq) and TPAP (5 mol%). The reaction mixture was stirred at rt for 14 h (reaction progress controlled by TLC). Most of the solvent was evaporated under reduced pressure and

-unsaturated aldehyde 15 was isolated after purification by column chromatography.

(E)-2-methyl-3-(thiophen-2-yl)prop-2-en-1-ol (15g)

The title compound was prepared from ester 13g (550 mg, 2.80 mmol) following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 3:1) afforded 15g as

yellow oil (175 mg, 1.15 mmol, 41% over two steps). 1H NMR (400.0 MHz, CDCl3): ppm 9.52 (s, 1H), 7.60 (d, 1H), 7.42 (s, 1H), 7.39 (d, 1H) 7.17 (dd, 1H), 2.09 (s, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 197.2 (CH), 144.8 (CH), 141.8 (C), 138.0 (C), 135.7 (CH), 134.1 (CH), 130.7 (CH), 13.7 (CH3). HRMS was not measured.

(E)-2-methyl-3-(m-tolyl)prop-2-en-1-ol (15h)

The title compound was prepared from ester 13h (300 mg, 1.47 mmol) following general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1) afforded 15h as yellow oil (183 mg, 1.14 mmol, 77% over two steps). 1H NMR (400.0 MHz, CDCl3): ppm

9.58 (s, 1H), 7.35 (d, 1H), 7.34 (d, 2H), 7.24-7.20 (m, 2H), 2.06 (s, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 195.6 (CH), 150.1 (CH), 138.4 (C), 138.2 (C), 135.1 (C), 130.7

O

S

O

A direct access to secondary alcohols

85

(CH), 130.4 (CH), 128.6 (CH), 127.1 (CH), 21.4 (CH3), 11.0 (CH3). HRMS was not measured.

(S,E)-2,5-dimethyl-1-(thiophen-2-yl)hex-1-en-3-ol (16g)

The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OD-H column, n-heptane/i-PrOH 99:1, 40 °C isotherm, detection at 274 nm, retention time (min): 39.2 (major) and 45.5 (minor).

(S,E)-2,5-dimethyl-1-(m-tolyl)hex-1-en-3-ol (16h)

The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OD-H column, n-heptane/i-PrOH 99:1, 40 °C isotherm, detection at 209 nm, retention time (min): 21.7 (major) and 24.8 (minor).

4.6 References

1. R. Noyori and S. Hashiguchi, in Applied Homogeneous Catalysis with Organometallic Compounds, eds. B. Cornils and W. A. Herrmann, Wiley-VCH Verlag GmbH, 2008, 552–571.

2. M. R. Luderer, W. F. Bailey, M. R. Luderer, J. D. Fair, R. J. Dancer, and M. B. Sommer, Tetrahedron Asymmetry, 2009, 20, 981–998.

3. T. Mukaiyama, K. Soai, T. Sato, H. Shimizu, and K. Suzuki, J. Am. Chem. Soc., 1979, 101, 1455–1460.

4. B. Weber and D. Seebach, Tetrahedron, 1994, 50, 6117–6128.

5. N. Oguni and T. Omi, Tetrahedron Lett., 1984, 25, 2823–2824.

6. M. Kitamura, S. Suga, K. Kawai, and R. Noyori, J. Am. Chem. Soc., 1986, 108, 6071–6072.

7. D. Seebach, L. Behrendt, and D. Felix, Angew. Chem. Int. Ed., 1991, 30, 1008–1009.

8. B. Weber and D. Seebach, Tetrahedron, 1994, 50, 7473–7484.

9. Y. Muramatsu and T. Harada, Angew. Chem. Int. Ed., 2008, 47, 1088–1090.

10. Y. Muramatsu, S. Kanehira, M. Tanigawa, Y. Miyawaki, and T. Harada, Bull. Chem. Soc. Jpn., 2010, 83, 19–32.

OH

S

OH

Chapter 4

86

11. C.-S. Da, J.-R. Wang, X.-G. Yin, X.-Y. Fan, Y. Liu, and S.-L. Yu, Org. Lett., 2009, 11, 5578–5581.

12. Y. Liu, C.-S. Da, S.-L. Yu, X.-G. Yin, J.-R. Wang, X.-Y. Fan, W.-P. Li, and R. Wang, J. Org. Chem., 2010, 75, 6869–6878.

13. E. Fernández-Mateos, B. Maciá, D. J. Ramón, and M. Yus, Eur. J. Org. Chem., 2011, 2011, 6851–6855.

14. E. Fernández-Mateos, B. Maciá, and M. Yus, Adv. Synth. Catal., 2013, 1249–1254.

15. A. V. R. Madduri, A. J. Minnaard, and S. R. Harutyunyan, Chem. Commun., 2012, 48, 1478–1480.

16. A. V. R. Madduri, S. R. Harutyunyan, and A. J. Minnaard, Org. Biomol. Chem., 2012, 10, 2878–2884.

17. A. V. R. Madduri, Total syntheses of (-)-Borrelidin and (-)-Doliculide and the development of the catalytic asymmetric addition of Grignard reagents to ketones, 2012.

18. S. H. Bertz, S. Cope, M. Murphy, C. A. Ogle, and B. J. Taylor, J. Am. Chem. Soc., 2007, 129, 7208–7209.

19. S. H. Bertz, R. A. Hardin, M. D. Murphy, C. A. Ogle, J. D. Richter, and A. A. Thomas, Angew. Chem. Int. Ed., 2012, 51, 2681–2685.

Enantioselective copper(I)-catalyzed alkylation of aryl alkyl ketones with Grignard reagents

A further application of a new catalytic system, the combination of copper(I) and (S,RFe)-reverse Josiphos, in the 1,2-addition of Grignard reagents to aryl alkyl ketones is presented. A study of the influence of the substitution pattern of the phenyl ring on the enantioselectivity showed, that substitution at the meta position is preferred. The limit of the reaction in terms of reactivity was found, when biaryl ketones were used as starting material.

Parts of this chapter will be published:

M. Hanstein, S.R. Harutyunyan, A.J. Minnaard, manuscript in preparation.

Chapter 5

88

5.1 Introduction

Tertiary alcohols are a general motif found in many natural products, pharmaceuticals and crop protection products. This functional group is challenging to synthesize. One way to form tertiary alcohols is the asymmetric 1,2-addition of organometallic reagents to ketones, so forming a new C-C bond. This stereoselective addition comes with two problems: first the low electrophilicity of the carbonyl group influences the reactivity, and second the steric similarity of substituents surrounding the carbonyl group makes the discrimination between the two faces of the double bond difficult.1 For these reasons, there are not much efficient catalyst systems reported promoting the addition with high enantioselectivities, as for the corresponding reaction with aldehydes (see for more details chapter 4). Therefore, the development of new catalytic systems creating such stereocenters in an asymmetric fashion, are of great importance.2,3

Fu and co-workers were the first to report this reaction in an asymmetric manner.4 Their initial study, the addition of diphenylzinc to aryl alkyl ketones, was based on the same catalyst Noyori previously described for the same reaction using aldehydes as substrates (Scheme 5.1).5

R1 R2

OPh2Zn+(3.5 eq)

15 mol% (+)-DAIB1.5 eq MeOHtoluene, rt, 48 h

R1 R2

HO Ph

53 to 91% yield60 to 91% ee

Me2N

HOMeMe

Me

(+)-DAIB

*

Scheme 5.1. Enantioselective addition of Ph2Zn to ketones catalyzed by (+)-DAIB.

A further development of this reaction was reported by Yus and co-workers, using a mixture of catalytic amounts of a camphorsulfonamide derivative L1 and an excess of Ti(OiPr)4. With this catalytic system, depending on the substrate, yields ranging between 25-95% and enantioselectivities up to 89% were obtained (Scheme 5.2).6

Enantioselective copper(I)-catalyzed alkylation of aryl alkyl ketones

89

R1 R2

OEt2Zn+(2.4 eq)

20 mol% L1toluene, rt, 4-14 d R1 R2

HO Et

25 to 95% yield13 to 89% ee

Ti(OiPr)4+(1.3 eq)

OH

O2S NH

L1

*

Scheme 5.2. Enantioselective addition of diethylzinc to ketones catalyzed by L1 and Ti(OiPr)4.

Later, the groups of Walsh and Yus developed a catalytic system based on bis(sulfonamide) ligand L2 and Ti(OiPr)4. Ligand L2 made not only addition of dialkylzinc reagents7–9 possible, but also diphenylzinc,10,11 divinylzinc12 and functionalized dialkylzinc reagents could be used. This made the reaction broadly applicable. In addition, the reaction time was reduced to 6-120 h depending on the substrate and zinc reagent (Scheme 5.3). In most cases, aryl alkyl ketones and substituted aryl alkyl ketones were used as substrates, giving tertiary alcohols with yields around 68-99% and enantioselectivities in the range of 46-99%.

R1 R2

OEt2Zn+(1.6 eq)

10 mol% L2toluene, rt, 6-120 h R1 R2

HO Et

68 to 99% yield46 to 99% ee

Ti(OiPr)4+(1.2 eq) NH

O2S

NHO2S

OH

OH

L2

*

Scheme 5.3. Enantioselective addition of zinc reagents to aryl alkyl ketones promoted by ligand L2.

In the catalytic systems shown up to now, the major drawback is the application of stoichiometric amounts of Ti(OiPr)4. Therefore, our group was interested to investigate a new catalytic system, in which the amount of the catalytic species is reduced and still the components of the catalytic system are commercially available or easy accessible. A first example of an enantioselective catalytic addition of

-unsaturated ketones was reported using a combination of 2 and (S,RFe)-reverse Josiphos L3 (Scheme 5.4).13,14

Chapter 5

90

R1 R2

O

X

R3 MgBr+1.3 eq

5 mol% CuBr.SMe26 mol% L3MTBE, 78 °C, 12 h R1 R2

HO

X

R3

R1: Ar, Alk; R2: Alk; X : Me, Br

R3: Alk yields up to 96%ee's up to 98%

FePP

H3C H

rev-JosiPhos (S, RFe)-L3

Scheme 5.4. Asymmetric addition of Grignard reagents to -unsaturated -substituted ketones.

Further investigations showed that the same catalytic system can be applied in the reaction with aryl alkyl ketones, to give access to a wide range of benzylic tertiary alcohols.15 Most of the ketones used in this study are based on an acetophenone scaffold.

5.2 Goal of this study

The goal of the present study is to establish, whether ketones possessing a longer alkyl chain undergo asymmetric copper(I)-catalyzed 1,2-addition with Grignard reagents with high enantioselectivities and yields. In addition, we hope to find a clear correlation between steric and electronic effects of the aryl substituents on the enantioselectivity. Due to the results of the previous studies, the commercially available iso-butylmagnesium bromide was chosen for this investigation, since the enantioselectivity of product 2a is comparable to 2b, and high (Scheme 5.5).15

O5 mol% CuBr.SMe2

6 mol% L3MTBE, 78 °C, 14 h

CF3

F3C

CF3

F3CR3 MgBr+(1.3 eq)

a R3: iBub R3: 2-ethyl butyl

2a: 95% yield, 95% ee 2b: 95% yield, 98% ee

1

HO R3

Scheme 5.5. Asymmetric addition to 3,5-difluoromethyl acetophenone.

Enantioselective copper(I)-catalyzed alkylation of aryl alkyl ketones

91

5.3 Results and discussion

In the first set of reactions, the optimized reaction conditions were applied to aryl alkyl ketones possessing an electron withdrawing substituent in various positions. Here, we tried to map the influence of the steric and electronic effects of the substrate on the enantioselectivity. In the first three examples, an acetophenone scaffold was used (Table 5.1, entries 1-3). It turned out that the meta substituted product gave the highest ee (4b, 64%) among the three fluorinated substrates.

A similar observation was made when brominated phenones with a longer alkyl chain were used. The meta substituted product showed the highest ee (4e, 80%) in this set of substrates (entries 4-6). The two substrates 3d and 3g with the bromo substituent in the ortho position, showed only around 10% conversion, and separation of the starting material from the product by column chromatography was not possible. Therefore, no product was isolated for further analysis to determine the ee.

Table 5.1. Aryl alkyl ketones with electron withdrawing substituent on the aromatic ring.

5 mol% CuBr.SMe26 mol% L3

1.3 eq iBuMgBrMTBE, 78 °C, 14 h

R2

O

R2

HO

+ R2

OH

3 4 5

R1 R1 R1

entry Substrate yielda eeb ratio [%]c

3:4:5

1 3a O

F 4a 36% rac 11:42:47

2 3b

O

F

4b 37% 64% 12:67:21

3 3c O

F 4c 24% 27% 9:79:12

4 3d O

Br 4d n.d. n.d. 17:11:71d

Chapter 5

92

entry Substrate yielda eeb ratio [%]c

3:4:5

5 3e

O

Br

4e 47% 80% 20:50:30

6 3f O

Br 4f 50% 61% 15:68:17

7 3g O

Br 4g n.d. n.d. 30:12:58d

8 3h O

Br 4h 32% 71% 22:70:8

9 3he

O

Br 4h 17% n.d. 14:52:34

a The isolated yield was obtained after column chromatography. b The enantiomeric excess was determined by HPLC; see experimentals. c The ratio was determined by 1H NMR spectroscopy of the crude product. d The approximate ratio was determined from the APT spectrum, due to overlapping signals in the 1H NMR spectrum. e Standard conditions, but the reaction time was extended to three days.

It is remarkable, that using iso-butylmagnesium bromide as Grignard reagent the meta position of the phenyl ring shows to be favored in terms of enantioselectivity. For the substrates possessing an ortho substituent, a decrease in conversion was observed. In general the ketones used in this study show a lower conversion in comparison to the substrates based on the acetophenone scaffold.15

Since the results of this study differ from the results obtained with the acetophenone derivatives, mostly in terms of yield, the influence of the change from 2-ethylbutylmagnesium bromide to iso-butylmagnesium bromide was investigated by performing blank reactions at -78 °C with substrate 1 and both Grignard reagents.15 The distribution between 1,2-addition product, 1,2-reduction product and starting material was determined by NMR, but no significant difference between the distribution of the compounds was found. Therefore, the increase of the amount of 1,2-reduction product cannot be explained by the change of the Grignard reagent. Instead, the change to ketones with longer alkyl chains, is accompanied with a higher tendency to enolize, which probably causes the lower conversion. To study the possibility that the reactions were just slower, the reaction time was extended to three

Enantioselective copper(I)-catalyzed alkylation of aryl alkyl ketones

93

days. This was tested with two different substrates (Table 5.1, entry 9; Table 5.2, entries 5). Only for substrate 3l a significant change in conversion was observed, combined with an increase in enantioselectivity. For the other substrate 3h more 1,2-reduction product was formed.

Next, unsubstituted aryl alkyl ketones and acetylfuranes were investigated in terms of conversion and enantioselectivity (Table 5.2). The conversions and isolated yields found are similar to the ones of the substituted aryl ketones, but the enantioselectivity decreased and ranged between 37-65%. Comparing enantioselectivities of products 4i, 4j and 4l it was noticed, that product 4j shows the highest ee among the three products. This result can be explained by the difference in steric bulk between a butyl group and a benzyl group. The ee might also be influenced by -intercations between the phenyl ring of the benzyl group with the copper. The low conversion of 3l to the desired product is caused by the stable enol-tautomer, which is preferred due to conjugation in the system.

The addition of iso-butyl Grignard reagent to substrate 3k would give a non-chiral compound, therefore ethylmagnesium bromide was added to give the same tertiary alcohol as obtained in entry 1 (Table 5.2, entries 1 and 3). The reaction with ethylmagnesium bromide (entry 3), went to full conversion since in the crude 1H NMR no signals were detected belonging to the starting material. In addition, there was no formation of the 1,2-reduction by-product noticed.

The two furans 3m and 3n both reacted with the iso-butylmagnesium bromide (entries 6 and 7), but product 4m was found as a racemic compound whereas 4n shows a relatively high ee of 65% (being in the same range as the substituted phenones).

Chapter 5

94

Table 5.2. Asymmetric 1,2-addition to unsubstituted alkyl aryl and heteroaryl ketones.

5 mol% CuBr.SMe26 mol% L3

1.3 eq iBuMgBrMTBE, 78 °C, 14 h

R2

O

R2

HO

+ R2

OH

3 4 5

entry Substrate yielda eeb ratio [%]c

3:4:5

1 3i O

4i 39% 48% 16:62:22d

2 3j O

4j 32% 56% 19:66:15d

3 3k O

4k 67%e 37% 0:100:0

4 3l O

4l 12% 40% 35:29:36

5f 3l

O

4l 19% 52% 6:56:38

6 3m O O

4m 35% Rac 37:55:08

7 3n O O

4n 21% 65% 51:38:11

a The isolated yield was obtained after column chromatography. b The enantiomeric excess was determined by HPLC. c The ratio was determined by 1H NMR spectroscopy of the crude product. d The approximate ratio was determined from the APT spectrum. e Ethylmagnesium bromide was added. f Standard condition, but the reaction time was extended to three days.

Additionally, ketones with more steric hindrance and biaryl ketones were studied as well (Scheme 5.6). The two ketones 3o and 3p showed some conversion, but too little to isolate the products and determine the ee. For the substrates 3q and 3r no reaction was observed. This indicates, that an increase of the steric bulk next to the C=O double bond prevents the reaction. An exception was substrate 3s, which gave selectively the 1,2-reduction product. The two biaryl ketones 3t and 3u were investigated, to see if chiral induction is possible with similar substituents next to the carbonyl double bond. However, both substrates showed no reactivity at all. To

Enantioselective copper(I)-catalyzed alkylation of aryl alkyl ketones

95

broaden the scope of the reaction using heteroaromatic substrates, ketone 3v was tested in the reaction as an example, but no product formation was observed.

OMe OBr

O

NO

O

CF3

O

O O

3o 3p 3q 3r

3s 3t 3u 3v

Scheme 5.6. Bulky substrates in the 1,2-addition.

5.4 Conclusion

The results of this study show, that several aryl alkyl ketones with alkyl chains longer than methyl undergo the 1,2-addition using our new catalytic system. However, enolization and 1,2-reduction become considerable competitors for the desired 1,2-addition. In addition to acetylthiophene, also acetylfurans showed desired reactivity in the 1,2-addition.

In general, for phenones, aryl alkyl ketones with an unsubstituted phenyl ring, a decrease in enantioselectivity was observed and in addition the apparent conversion was far from complete after the standard reaction time. During a longer reaction time no significant increase of ee was observed. Remarkably, for ketone 3h a bit more side product was formed, whereas for ketone 3l more starting material was converted to the desired product. This does not match with enolization of the starting material, as one would expect the enolate to be stable under the reaction conditions.

The results of reactions with substituted aryl alkyl ketones showed, that the substitution pattern on the phenyl ring has an influence on the enantioselectivity. Substrate 3a, with a fluoro substituent in the ortho position, produces a racemate, and the substrates 3c and 3e possessing a substituent in meta position showed the highest ee’s.

A limitation in terms of substrate scope was found, when biaryl ketones and bulkier alkyl groups were used. Here, 1,2-addition was hardly observed. In the future more

Chapter 5

96

ligands need to be screened to adjust the catalytic system in such a way, that addition of linear Grignard reagents to aryl alkyl ketones will be possible. This means the bulkiness of the ligand will influence the differentiation of the two faces of the carbonyl double bond.

5.5 Experimental

For general information see experimental of chapter 2.

5.5.1 Synthesis of the ketones 3d, 3g

General procedure for the synthesis of secondary alcohols S1a, b16

O

Br

alkylMgBr,THF0 °C to rt, on

alkyl

OH

Br 2-bromobenzaldehyde (1 eq) was dissolved in THF (0.8 mL/mmol) and cooled with an ice bath to 0 °C. To this the Grignard reagent (1.7 eq) was added dropwise. The ice bath was removed and the reaction mixture was stirred at rt for 16 h. The reaction was quenched by adding saturated aq. NH4Cl. The layers were separated and the water layer was extracted three times with diethyl ether. The combined organic layers were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. Purification by column chromatography afforded S1a and S1b as colorless oil. For S1a the Grignard reagent was purchased from Aldrich as a 3 M solution in diethyl ether. For S1b the Grignard reagent was prepared as follow: magnesium (2.43 g, 100 mmol, 2.5 eq) was activated with an iodine crystal and 5 mL diethyl ether was added. To this was added dropwise a solution of 1-bromopropane (4.91 g, 40 mmol, 1.7 eq) in 10 mL diethyl ether to keep the reaction mixture refluxing. The reaction mixture was stirred for 18 h at rt before further use.

1-(2-bromophenyl)propan-1-ol (S1a)

Following the general procedure, 2.40 g (13 mmol) 2-bromobenzaldehyde was reacted with 11 mL ethylmagnesium bromide to give 1.65 g S1a (7.7 mmol, 59%) after column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.35 in pentane/diethyl ether, 9:1). 1H

NMR (400.0 MHz, CDCl3): ppm 7.51 (m, 2H), 7.32 (m, 1H), 7.21 (m, 1H), 5.00 (q,

OH

Br

Enantioselective copper(I)-catalyzed alkylation of aryl alkyl ketones

97

1H), 2.25 (s, 1H), 1.82 (m, 1H), 1.68 (m, 1H), 0.99 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 143.5 (C), 132.5 (CH), 128.6 (CH), 127.6 (CH), 127.3 (CH), 122.1 (C), 74.1 (CH), 30.5 (CH2), 10.0 (CH3). HRMS (APCI+): m/z [M-H2O]+ calcd for C9H10Br: 196.9960; found: 196.9959.

1-(2-bromophenyl)butan-1-ol (S1b)

Following the general procedure, 4.35 g (23 mmol) 2-bromobenzaldehyde was reacted with 15 mL propylmagnesium bromide to give 2.79 g S1b (12 mmol, 52%) after column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.46 in

pentane/diethyl ether, 9:1). 1H NMR (400.0 MHz, CDCl3): ppm 7.51 (m, 2H), 7.31 (m, 1H), 7.10 (m, 1H), 5.06 (m, 2H), 1.68 (m, 2H), 1.53-1.39 (m, 2H), 0.96 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 143.9 (C), 132.6 (CH), 128.6 (CH), 127.7 (CH), 127.3 (CH), 122.0 (C), 72.6 (CH), 39.8 (CH2), 19.0 (CH2), 13.9 (CH3). HRMS (APCI+): m/z [M-H2O]+ calcd for C10H12OBr: 211.0017; found: 211.0117.

Oxidation to ketones 3d, 3g17

1-(2-bromophenyl)propan-1-one (3d)

OH

Br

DMP, NaHCO3, CH2Cl20 °C for addition, rt 2 h O

Br

S1a (724 mg, 3.4 mmol, 1 eq) was dissolved in dry CH2Cl2. At 0 °C (ice bath) Dess-Martin periodinane (2.0 g, 4.7 mmol, 1.4 eq) and NaHCO3 (566 mg, 6.7 mmol, 2 eq) were added. The reaction was subsequently stirred at rt for 2 h (until TLC indicated consumption of the starting material). The excess of oxidant was destroyed by adding 2-propanol (2 mL), the mixture was poured into 100 mL pentane/diethyl ether (9:1) (Rf = 0.60 pentane/diethyl ether, 9:1) and allowed to stand for 30 min at rt. The mixture was directly loaded on to a silica column. After purification 3d (722 mg, 3.4 mmol, 100%) was isolated. 1H NMR (400.0 MHz, CDCl3): ppm 7.69 (d, 1H), 7.36 (d, 1H), 7.28 (m, 2H), 2.93 (q, 2H), 1.21 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 205.0 (C), 142.0 (C), 133.5 (CH), 131.5 (CH), 128.1 (CH), 127.4 (CH), 118.5 (C), 36.0 (CH2), 8.1 (CH3). HRMS (ESI+): m/z [M+H] calcd for C9H10OBr: 212.9909; found: 212.9910.

OH

Br

Chapter 5

98

1-(2-bromophenyl)butan-1-one (3g)

OH

Br

IBX, NaHCO3, CH2Cl20 °C for addition, rt 16 h O

Br

S1b (2.00 g, 8.7 mmol) was dissolved in dry CH2Cl2. At 0 °C (ice bath) 2-iodoxybenzoic acid (6.25 g, 14.7 mmol, 1.7 eq) and NaHCO3 (1.47 g, 17.5 mmol, 2 eq) were added. The reaction was then stirred at rt for 16 h (until TLC indicated consumption of the starting material). The solvent of the reaction mixture was evaporated. The crude product was purified by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.33 in pentane/diethyl ether, 9:1) to give 3g as colorless oil (0.75 g, 3.3 mmol, 38%). 1H NMR (400.0 MHz, CDCl3): ppm 7.69 (d, 1H), 7.36 (d, 1H), 7.28 (m, 2H), 2.93 (q, 2H), 1.21 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 204.5 (C), 142.1 (C), 133.6 (CH), 131.3 (CH), 128.2 (CH), 127.3 (CH), 118.5 (C), 44.6 (CH2), 17.6 (CH2), 13.7 (CH3). HRMS (ESI ): m/z [M+H] calcd for C10H12OBr: 227.0066; found: 227.0066.

5.5.3 General procedure for the copper-catalyzed 1,2-addition to ketones

2 (0.015 mmol, 5 mol%) and (S,RFe)-reverse Josiphos (L3) (0.018 mmol, 6 mol%) were dissolved in dry MTBE (2 mL) and stirred at rt for 15 min. To this mixture was added the corresponding ketone (0.3 mmol, 1 eq) in 2 mL dry MTBE. Then the mixture was cooled to 78 °C. After stirring for 15 min at 78 °C the corresponding Grignard reagent (0.36 mmol, 1.2 eq) was added within 30 min. The reaction mixture was stirred at 78 °C for 14 h. The reaction was quenched with 1 mL MeOH and 2 mL saturated aq. NH4Cl. After the reaction mixture reached rt, the layers were separated and the water layer was extracted three times with CH2Cl2. The combined organic layers were dried over MgSO4, filtered and the filtrate was concentrated in vacuo. From the crude product a sample was taken for GC-MS and NMR analysis to determine the ratio of the 1,2-addition product, reduction product and starting material. The enantiomeric ratio was determined after purification by column chromatography.

Enantioselective copper(I)-catalyzed alkylation of aryl alkyl ketones

99

2-(2-fluorophenyl)-4-methylpentan-2-ol (4a)

The title compound was prepared from ketone 3a following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.37 in pentane/diethyl ether, 9:1) afforded 4a as a light yellow oil (21.6 mg, 0.110 mmol, 36%, rac). 1H

NMR (400.0 MHz, CDCl3): ppm 7.29 (m, 1H), 7.19 (m, 1H), 7.16 (m, 1H), 6.92 (m, 1H), 3.48 (m, 1H), 1.74 (m, 2H), 1.61 (m, 1H), 1.54 (s, 3H), 0.89 (d, 3H), 0.75 (d, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 164.1 (C), 151.1 (C), 129.5 (CH), 121.4 (CH), 113.1 (CH), 112.1 (CH), 75.0 (C), 52.6 (CH2), 31.4 (CH3), 24.6 (CH), 24.6 (CH3), 24.3 (CH3). Optical rotation [ D20= -5.7 (c = 0.1, CHCl3). HRMS (APCI+): m/z [M HF]+ calcd for C12H17O: 177.1274; found: 177.1269. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel AD-H column, n-heptane/i-PrOH 99:1, 40 °C isotherm, detection at 206 nm, retention time (min): 14.9 and 15.4.

2-(3-fluorophenyl)-4-methylpentan-2-ol (4b)

The title compound was prepared from ketone 3b following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.43 in pentane/diethyl ether, 9:1) afforded 4b as a light yellow oil (21.6 mg, 0.110 mmol, 36%, 64% ee).

1H NMR (400.0 MHz, CDCl3): ppm 7.31 (m, 1H), 7.19 (m, 1H), 7.16 (m, 1H), 6.92 (m, 1H), 3.48 (m, 1H), 1.77 (m, 2H), 1.61 (m, 1H), 1.54 (s, 3H), 0.89 (d, 3H), 0.75 (d, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 164.1 (C), 151.1 (C), 129.5 (CH), 121.4 (CH), 113.1 (CH), 112.1 (CH), 75.0 (C), 52.6 (CH2), 31.4 (CH3), 24.6 (CH), 24.6 (CH3), 24.3 (CH3). Optical rotation [ D20= -1.2 (c = 0.9, CHCl3). HRMS (APCI+): m/z [M HF] calcd for C12H17O: 177.1274; found: 177.1275. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel AS-H column, n-heptane/i-PrOH 100:0, 40 °C isotherm, detection at 207 nm, retention time (min): 16.8 (major) and 18.3 (minor).

2-(4-fluorophenyl)-4-methylpentan-2-ol (4c)

The title compound was prepared from ketone 3c following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.53 in pentane/diethyl ether, 9:1) afforded 4c as a light yellow oil (13.9 mg, 0.071 mmol, 24%,

32% ee). 1H NMR (400.0 MHz, CDCl3): ppm 7.40 (q, 2H), 7.01 (t, 2H), 1.74 (m, 2H), 1.60 (m, 2H), 1.54 (s, 3H), 0.88 (d, 3H), 0.76 (d, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 162.7 (C), 144.0 (C), 126.5 (CH), 126.4 (CH), 114.8 (CH), 114.6 (CH), 75.0 (C), 52.9 (CH2), 31.5 (CH3), 30.3 (CH), 24.4 (CH3), 24.3 (CH3). Optical rotation [ D20= -6.3 (c =

F

HO

HO

F

HO

F

Chapter 5

100

0.2, CHCl3). HRMS (APCI+): m/z [M HF] calcd for C12H17O: 177.1273; found: 177.1274. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel AD-H column, n-heptane/i-PrOH 99:1, 40 °C isotherm, detection at 207 nm, retention time (min): 22.2 (major) and 23.1 (minor).

3-(3-bromophenyl)-5-methylhexan-3-ol (4e)

The title compound was prepared from ketone 3e following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.49 in pentane/diethyl ether, 9:1) afforded 4e as a light yellow oil (38.2 mg, 0.141 mmol, 47%, 80% ee). 1H NMR (400.0 MHz, CDCl3): ppm 7.56 (m, 1H), 7.35 (m, 1H), 7.28

(m, 1H), 7.19 (m, 1H), 1.71 (m, 2H), 1.59 (m, 2H), 1.54 (s, 1H), 1.49 (m, 1H), 0.91 (d, 3H), 0.74 (d, 3H), 0.71 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 148.8 (C), 129.5 (CH), 129.2 (CH), 128.7 (CH), 124.1 (CH), 122.4 (C), 77.5 (C), 51.4 (CH2), 36.2 (CH2), 24.5 (CH3), 24.3 (CH3), 24.1 (CH), 7.5 (CH3). Optical rotation [ D20= -2.4 (c = 1.0, CHCl3). HRMS (ESI ): m/z [M-H]+ calcd for C13H18OBr: 271.0515; found: 271.0511. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel AD-H column, n-heptane/i-PrOH 99:1, 40 °C isotherm, detection at 203 nm, retention time (min): 12.48 (major) and 13.01 (minor).

3-(4-bromophenyl)-5-methylhexan-3-ol (4f)

The title compound was prepared from ketone 3f following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.54 in pentane/diethyl ether, 9:1) afforded 4f as a light yellow oil (41.0 mg, 0.151 mmol,

50%, 61% ee). 1H NMR (400.0 MHz, CDCl3): ppm 7.37 (dd, 2H), 7.19 (dd, 2H), 1.71 (m, 2H), 1.58 (m, 1H), 1.51 (s, 1H), 1.46 (m, 2H), 0.83 (d, 3H), 0.66 (d, 3H), 0.63 (t, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 145.2 (C), 131.0 (2 x CH), 127.4 (2 x CH), 120.0 (C), 77.5 (C), 51.4 (CH2), 36.7 (CH2), 24.6 (CH3), 24.3 (CH3), 24.1 (CH), 7.5 (CH3). Optical rotation [ D20= -1.7 (c = 1.0, CHCl3). HRMS (ESI ): m/z [M-H]+ calcd for C13H18OBr: 271.0515; found: 271.0510. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OD-H column, n-heptane/i-PrOH 99:1, 40 °C, detection at 222 nm, retention time (min): 12.59 (major) and 13.41 (minor).

HO

Br

HO

Br

Enantioselective copper(I)-catalyzed alkylation of aryl alkyl ketones

101

2-(4-bromophenyl)-4-methyl-1-phenylpentan-2-ol (4h)

The title compound was prepared from ketone 3h following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.46 in pentane/diethyl ether, 9:1) afforded 4h as a light yellow oil (32.0 mg, 0.096 mmol, 32%, 71% ee). 1H NMR (400.0 MHz,

CDCl3): ppm 7.42 (d, 2H), 7.19 (m, 5H), 6.90 (m, 2H), 3.10 (d, 1H), 2.99 (d, 1H), 1.94 (dd, 1H), 1.68 (dd, 1H), 1.52 (m, 2H), 0.88 (d, 3H), 0.69 (d, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 145.1 (C), 135.2(C), 130.9 (2 x CH), 130.6 (2 x CH), 128.2 (2 x CH), 127.5 (2 x CH), 126.8 (CH) 120.2 (C), 77.2 (C), 50.8 (CH2), 50.7 (CH2), 24.6 (CH3), 24.3 (CH3), 24.0 (CH). Optical rotation [ D20= -47.9 (c = 1.1, CHCl3). HRMS (APCI+): m/z [M-OH]+ calcd for C18H20Br: 315.0743; found: 315.0732. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OD-H column, n-heptane/i-PrOH 99:1, 40 °C, detection at 190 nm, retention time (min): 16.17 (major) and 17.30 (minor).

5-methyl-3-phenylhexan-3-ol (4i)

The title compound was prepared from ketone 3i following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.28 in pentane/diethyl ether, 9:1) afforded 4i as a light yellow oil (22.6 mg, 0.117 mmol, 39%, 48% ee).

1H NMR (400.0 MHz, CDCl3): ppm 7.39-7.30 (m, 4H), 7.21 (m, 1H), 1.82 (m, 2H), 1.68 (m, 2H), 1.54 (m, 1H), 1.43 (s, 1H), 0.90 (d, 3H), 0.73 (t, 3H) 0.68 (d, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 146.1 (C), 127.9 (2 x CH), 126.1 (CH), 125.4 (2 x CH), 77.7 (C), 51.4 (CH2), 36.6 (CH2), 24.5 (CH3), 24.4 (CH3), 24.1 (CH), 7.6 (CH3). Optical rotation [ D20= 3.3 (c = 1.1, CHCl3). HRMS-ESI+: m/z [M]+ calcd for C13H18OBr: 270.0619; found: 271.0511. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OD-H column, n-heptane/ i-PrOH 100:0, 40 °C, detection at 205 nm, retention time (min): 20.97 (major) and 23.34 (minor).

2-methyl-4-phenylheptan-4-ol (4j)

The title compound was prepared from ketone 3j following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.31 in pentane/diethyl ether, 9:1) afforded 4j as a light yellow oil (20.1 mg, 0.097 mmol, 32%, 56% ee). 1H NMR (400.0 MHz, CDCl3): ppm 7.39-7.30 (m, 4H), 7.21 (m, 1H),

1.80 (m, 2H), 1.64 (m, 2H), 1.57 (m, 1H), 1.43 (s, 1H), 1.27 (m, 2H), 0.90 (d, 3H), 0.83 (t, 3H), 0.69 (d, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 146.5 (C), 127.9 (2 x CH),

HO

Br

HO

HO

Chapter 5

102

126.1 (CH), 125.2 (2 x CH), 77.6 (C), 51.7 (CH2), 46.5 (CH2), 24.5 (CH3), 24.4 (CH3), 24.0 (CH), 16.6 (CH2), 14.4 (CH3). Optical rotation [ D20= 2.6 (c = 0.5, CHCl3). HRMS (ESI ): m/z [M H] calcd for C14H21O: 205.1586; found: 205.1587. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OD-H column, n-heptane/i-PrOH 100:0, 40 °C, detection at 208 nm, retention time (min): 20.34 (major) and 22.02 (minor).

5-methyl-3-phenylhexan-3-ol (4k)

The title compound was prepared from ketone 3k following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.51 in pentane/diethyl ether, 9:1) afforded 4k as a light yellow oil (39.2 mg, 0.204 mmol, 68%, 37% ee).

1H NMR (400.0 MHz, CDCl3): ppm 7.39-7.31 (m, 4H), 7.21 (m, 1H), 1.83 (m, 3H), 1.61 (m, 1H), 1.62 (s, 1H), 1.55 (m, 1H), 0.90 (d, 3H), 0.83 (t, 3H), 0.69 (d, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 146.1 (C), 127.9 (2 x CH), 126.1 (CH), 125.4 (2 x CH), 77.7 (C), 51.4 (CH2), 36.6 (CH2), 24.6 (CH3), 24.4 (CH3), 24.1 (CH), 7.6 (CH3). Optical rotation [ D20= 0.2 (c = 1.1, CHCl3). HRMS (ESI ): m/z [M H2O] calcd for C13H19: 175.1481; found: 175.1476. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel OD-H column, n-heptane/i-PrOH 99:1, 40 °C, detection at 209 nm, retention time (min): 11.96 (minor) and 12.48 (major).

4-methyl-1,2-diphenylpentan-2-ol (4l)

The title compound was prepared from ketone 3l following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.59 in pentane/diethyl ether, 9:1) afforded 4l as a light yellow oil (20.9 mg, 0.082 mmol, 27%,

40% ee). 1H NMR (400.0 MHz, CDCl3): ppm 7.31 (m, 4H), 7.23 (m, 1H), 7.18 (m, 3H), 6.90 (m, 2H), 3.14 (d, 1H), 3.01 (d, 1H), 1.97 (dd, 1H), 1.69 (dd, 1H), 1.47 (m, 2H), 0.80 (d, 3H), 0.61 (d, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 145.9 (C), 136.1 (C), 130.7 (2 x CH), 129.9 (CH), 129.0 (CH), 128.0 (2 x CH), 127.8 (2 x CH), 126.6 (CH), 126.2 (CH), 77.2 (C), 50.9 (CH2), 50.8 (CH2), 24.6 (CH3), 24.3 (CH3), 24.0 (CH). Optical rotation [ D20= -34.9 (c = 0.4, CHCl3). HRMS (APCI+): m/z [M-OH]+ calcd for C18H21: 237.1638; found: 237.1634. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralpak OJ-H column, n-heptane/i-PrOH 99:1, 40 °C, detection at 209 nm, retention time (min): 14.51 (minor) and 15.63 (major).

HO

HO

Enantioselective copper(I)-catalyzed alkylation of aryl alkyl ketones

103

2-(furan-2-yl)-4-methylpentan-2-ol (4m)

The title compound was prepared from ketone 3m following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.18 in pentane/diethyl ether, 9:1)

afforded 4m as a light yellow oil (17.7 mg, 0.105 mmol, 35%, rac). 1H NMR (400.0 MHz, CDCl3): ppm 7.34 (m, 1H), 6.30 (m, 1H), 6.19 (dd, 1), 1.91 (s, 1H), 1.78 (m, 2H), 1.64 (m, 1H), 1.55 (s, 3H), 0.83-0.79 (dd, 9H). 13C NMR (100.6 MHz, CDCl3): ppm 159.8 (C), 141.2 (CH), 110.1 (CH), 104.3 (CH), 71.8 (C), 50.3 (CH2), 27.3 (CH3), 24.5 (CH3), 24.1 (CH3), 23.9 (CH). HRMS (APCI+): m/z [M-HO]+ calcd for C10H15O: 151.1117; found: 151.1118.

4-methyl-2-(5-methylfuran-2-yl)pentan-2-ol (4n)

The title compound was prepared from ketone 3n following the general procedure. Purification by column chromatography (SiO2, pentane/diethyl ether 9:1, Rf = 0.68 in pentane/diethyl ether, 1:1)

afforded 4n as a light yellow oil (11.6 mg, 0.064 mmol, 21%, 65% ee). 1H NMR (400.0 MHz, CDCl3): ppm 6.04 (d, 1H), 5.87 (m, 1H), 2.27 (s, 3H), 1.76 (m, 2H), 1.64 (m, 1H), 1.53 (s, 3H), 0.86 (d, 3H), 0.80 (d, 3H). 13C NMR (100.6 MHz, CDCl3): ppm 158.0 (C), 150.8 (C), 105.9 (CH), 104.9 (CH), 71.6 (C), 50.2 (CH2), 27.0 (CH3), 24.5 (CH3), 24.3 (CH3), 23.9 (CH), 13.5 (CH3). Optical rotation [ D20= -1.4 (c = 0.6, CHCl3). HRMS (APCI+): m/z [M-OH]+ calcd for C11H17O: 165.1274; found: 165.1272. The enantiomeric ratio was determined by chiral HPLC analysis, Chiralcel AD-H column, n-heptane/i-PrOH 99:1, 40 °C, detection at 209 nm, retention time (min): 22.12 (minor) and 22.93 (major).

OHO

OHO

Chapter 5

104

5.6 References

1. O. Riant and J. Hannedouche, Org. Biomol. Chem., 2007, 5, 873–888.

2. D. J. Ramón and M. Yus, Angew. Chem. Int. Ed., 2004, 43, 284–287.

3. J. Christoffers and A. Baro, Eds., Quaternary Stereocenters: Challenges and Solutions for Organic Synthesis, 2006.

4. P. I. Dosa and G. C. Fu, J. Am. Chem. Soc., 1998, 120, 445–446.

5. M. Kitamura, S. Okada, S. Suga, and R. Noyori, J. Am. Chem. Soc., 1989, 111, 4028–4036.

6. D. J. Ramón and M. Yus, Tetrahedron, 1998, 54, 5651–5666.

7. M. Yus, D. J. Ramón, and O. Prieto, Tetrahedron Asymmetry, 2002, 13, 2291–2293.

8. C. García, L. K. LaRochelle, and P. J. Walsh, J. Am. Chem. Soc., 2002, 124, 10970–10971.

9. S.-J. Jeon, H. Li, C. García, L. K. LaRochelle, and P. J. Walsh, J. Org. Chem., 2005, 70, 448–455.

10. C. García and P. J. Walsh, Org. Lett., 2003, 5, 3641–3644.

11. O. Prieto, D. J. Ramón, and M. Yus, Tetrahedron Asymmetry, 2003, 14, 1955–1957.

12. H. Li, C. García, and P. J. Walsh, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 5425–5427.

13. A. V. R. Madduri, A. J. Minnaard, and S. R. Harutyunyan, Chem. Commun., 2012, 48, 1478–1480.

14. A. V. R. Madduri, A. J. Minnaard, and S. R. Harutyunyan, Org. Biomol. Chem., 2012, 10, 2878–2884.

15. A. V. R. Madduri, S. R. Harutyunyan, and A. J. Minnaard, Angew. Chem. Int. Ed., 2012, 51, 3164–3167.

16. T. Matsuda, M. Shigeno, M. Makino, and M. Murakami, Org. Lett., 2006, 8, 3379–3381.

17. H. M. Ge, L.-D. Zhang, R. X. Tan, and Z.-J. Yao, J. Am. Chem. Soc., 2012, 134, 12323–12325.

Summary

Summary

106

The work described in the first part of this thesis is about the synthesis of compounds, e.g. lipids and fatty acids, isolated from the cell wall of bacteria. The cell wall defines the inside of the cell from its surrounding, and functions as a protective barrier. A distinct example of a cell wall, with a low permeability, is the cell wall of Mycobacterium tuberculosis. Due to this characteristic of the cell wall, treatment with therapeutics is difficult. For a better insight into this problem, biochemists need isolated cell wall components to investigate this phenomenon. Since Mycobacterium tuberculosis is a slow growing bacterium and requires specific research facilities, the synthesis of those compounds by chemist provides an alternative.

Chemical synthesis has the advantage of a selective production of a specific cell wall component, with high regio- and stereoselectivity. Additionally, chemical synthesis leaves room for changes and adaptations to the requirements from a biochemical point, for example to have manipulating tools. The synthesis of such natural products is an important way to identify the chemical structure of the isolated compounds. This information is valuable for investigations at a molecular level.

In chapter 2 our efforts to synthesize the fatty acid, mycolic acid (Scheme 1), the main cell wall component of Mycobacterium tuberculosis, are described. A mycolic acid molecule can be divided into three parts: the mycolic chain, the meromycolate chain and a -hydroxy acid moiety. One option to assemble the meromycolate chain is to start from a common cyclopropyl building block. We envisioned to form the cyclopropyl ring in an asymmetric catalytic reaction, which would give a compound with two different functional groups to introduce the alkyl chains independently. One reaction type, which fulfills these requirements, is the intramolecular cyclopropanation of allylic diazoacetates reported by Doyle et al.. Following this route we faced the problem of reproducing reliable results of this reaction when carried out on a bigger scale. The synthesis of the -hydroxy acid moiety was investigated on a model substrate (A). The -keto ester A was converted in an asymmetric hydrogenation to the corresponding -hydroxy ester B with high yield (95%) and an excellent enantioselecitvity (>95%).

Summary

107

OH

COOH

mycolic chain

meromycolate chain

O

CHN2

O O O46%87% ee

0.1 mol% Rh2(4S-MEOX)4CH2Cl2, reflux, 12-18 h

Rh2(4S-MEOX)4

Rh Rh

N O

O N COOMe

HMeOOC

H

OH

MeOOC

O N COOMe

H

[(RuCl(T-BINAP))2(μ-Cl)3NH2Me2]EtOH, 20 bar, 50 °C, 25 h

95%>95% ee

O

O O

17O

OH O

17

A B

Scheme 1. Chemical structure of mycolic acid and key steps of the synthesis.

The synthesis of a second fatty acid, lactocacillic acid (Scheme 2), first isolated from the cell wall of Lactobacillus arabinosus, is explained in chapter 3. This fatty acid has like mycolic acid a cyclopropyl moiety as structural feature. Here, we followed a different strategy to introduce the cis-configured cyclopropyl ring. As a key intermediate we aimed to synthesize a chiral allylic alcohol with a cis-configured double bond. To achieve this we formed the chiral alcohol in a hetero asymmetric allylic alkylation, and built the cis-configured double bond in a ring-closing metathesis (RCM). The cyclopropyl ring was introduced in a Simmons-Smith reaction, where the chirality of the alcohol induced the chirality of the cyclopropyl ring. Unfortunately, the synthesis was not completed since all attempts to eliminate the alcohol group after the Simmons-Smith reaction failed.

Summary

108

OH

O

lactobacillic acid

O

O Br

5 mol% L1 5 mol% CuBr.SMe2

2eq C5H11MgBr 75 °C, CH2Cl2, 20 h O

O91%

97% ee

FePPh2

N H

(S,Rfe) Taniaphos L1

PPh2

O

O OOGrubbs II, toluene120 °C, 5 h

81%

Scheme 2. Chemical structure of lactobacillic acid and key steps of the synthesis.

The second part of this thesis is about a new methodology, lately developed in our institute. With this methodology chiral secondary and in particular tertiary alcohols can be synthesized in a single step by the copper(I)-catalyzed addition of Grignard reagents to either aldehydes or ketones. The development of new methodologies is essential in the field of chemistry in order to improve existing procedures as well as to develop novel routes for the synthesis of natural products. Nowadays the development of such a methodology requires not only regio- and stereospecificity, but also atom economy gets important to make chemical processes more sustainable.

In chapter 4 the copper(I)-catalyzed addition of Grignard reagents to aryl aldehydes -unsaturated aldehydes was investigated (Scheme 3). Initially the optimized

conditions, earlier reported for the 1,2-addition to acetophenones, were used. In comparison to the corresponding aryl ketones, for the aryl aldehydes only a moderate enantioselectivity and low isolated yields were observed. Several parameters were varied, also different additives were tested to suppress the fast blank reaction.

Summary

109

FePP

H3C H

rev-JosiPhos (S, RFe)-L2

OHO5 mol% CuBr.SMe2

6 mol% L21.3 eq Grignard

MTBE, 78 °C, 14 hH

R R

yield 6-48% ee 23-67%R: Me, F, CF3

Scheme 3. Copper(I)-catalyzed asymmetric alkylation of aldehydes with Grignard reagents.

In chapter 5 results of the asymmetric copper(I)-catalyzed 1,2-addition of Grignard reagents to aryl-alkyl ketones with longer alkyl chains are described. In contrast to the earlier investigation of aryl ketones based on the acetophenone scaffold, the ketones used in this study showed much lower enantioselectivities (Scheme 4). In addition the formation of a by-product, the 1,2-reduction product, was observed. A limit of the copper(I)-catalyzed alkylation with Grignard reagents was reached, when biaryl ketones and bulkier alkyl groups were used.

5 mol% CuBr.SMe26 mol% L2

1.3 eq iBuMgBrMTBE, 78 °C, 14 h

R2

O

R2

HO

+ R2

OH

R1 R1 R1

yield 12-50% ee 37-80%

by-productR1: F, BrR2: Et, Pr, Bz

Scheme 4. Copper(I)-catalyzed alkylation of aryl-alkyl ketones with Grignard reagents.

Summary

110

Samenvatting

Samenvatting

112

Het werk beschreven in het eerste gedeelte van dit proefschrift gaat over de synthese van verscheidene verbindingen, zoals lipiden en vetzuren, die zijn geïsoleerd uit de celwand van bacteriën. De celwand begrenst de celorganellen van de omgeving en functioneert als beschermende barrière. De celwand van Mycobacterium tuberculosis is een voorbeeld van een bijzondere celwand met een zeer lage doordringbaarheid, waardoor bestaande behandelingen niet zo effectief zijn. Om dit probleem verder te onderzoeken, is het van belang verschillende verbindingen uit de celwand met een hoge zuiverheid te verkrijgen. Omdat Mycobacterium tuberculosis één van de langzaamst groeiende bacteriën is en het kweken in zeer weinig laboratoria is toegestaan, is de chemische synthese van deze verbindingen een goede manier om deze verbindingen in handen te krjigen.

Het voordeel van chemische synthese is, dat de celwandcomponenten gericht gesynthetiseerd worden. Door middel van synthese kunnen de verbindingen worden verkregen met een hoge regio- en stereoselectiviteit. Bovendien biedt de chemische synthese ruimte voor aanpassingen, waardoor de synthese van derivaten van de celwandcomponenten mogelijk wordt. Met deze derivaten wordt het mogelijk de processen binnen de cel beter te volgen. Ook wordt het door de synthese van deze verbindingen mogelijk, hun chemische structuur volledig te identificeren. Deze gedetailleerde informatie is van belang, om de cellulaire processen op moleculair niveau te begrijpen.

In hoofdstuk 2 worden onze inspanningen gericht op de synthese van mycolzuur (Schema 1), een vetzuur uit de celwand van Mycobacterium tuberculosis, beschreven. Een molecuul mycolzuur bestaat uit drie delen: de mycol keten, de meromycol keten en een -hydroxycarbonzuur. De synthese van de twee cyclopropaanringen in de meromycol keten zou uitgaan van dezelfde bouwsteen. Deze bouwsteen bezit twee verschillende functionaliteiten, om de verschillende alkylstaarten aan de cyclopropaanring te koppelen. Een mogelijkheid voor het maken van deze bouwsteen is de intramoleculaire cyclopropanering van allylische diazoacetaten beschreven door Doyle et al.. Het opschalen van deze reactie bleek echter problematisch en in de toekomst moet voor dit deel van de synthese een nieuwe strategie worden ontworpen. De synthese van het -hydroxycarbonzuur is onderzocht aan de hand van een modelverbinding (A). Daarvoor werd ketoester A omgezet met een asymmetrische hydrogenering naar de hydroxyester B in een hoge opbrengst (95%) en een hoge enantiomere overmaat (>95%).

Samenvatting

113

OH

COOH

mycolic chain

meromycolic chain

O

CHN2

O O O46%87% ee

0.1 mol% Rh2(4S-MEOX)4CH2Cl2, reflux, 12-18 h

Rh2(4S-MEOX)4

Rh Rh

N O

O N COOMe

HMeOOC

H

OH

MeOOC

O N COOMe

H

[(RuCl(T-BINAP))2(μ-Cl)3NH2Me2]EtOH, 20 bar, 50 °C, 25 h

95%>95% ee

O

O O

17O

OH O

17

A B

Schema 1. De chemische structuur van mycolzuur en centrale stappen van de synthese.

In hoofdstuk 3 wordt een nieuwe syntheseroute voor het vetzuur, lactobacilluszuur uit de celwand van Lactobacillus arabinosus, gepresenteerd. Dit vetzuur heeft net als mycolzuur een cyclopropaanring als functionele groep. In deze syntheseroute zou de cis-configuratie van de cyclopropaanring worden gevormd met een Simmons-Smith reactie. Het uitgangsmateriaal hiervoor is een chiraal enantiomeerzuiver allylisch alcohol met een cis-geconfigureerde dubbele binding. Voor de synthese van het chirale allylische alcohol werd gebruik gemaakt van een hetero asymmetrische allylische alkylatie en vervolgens een ringsluitingsmetathese. Het chirale alcohol zou de stereoinformatie tijdens de Simmons-Smith reactie induceren, door middel van coördinatie tussen de hydroxyl groep en het zink. Uiteindelijk bleek het niet mogelijk het vetzuur volledig te synthetiseren, omdat de eliminatie van de hydroxylgroep niet succesvol was.

Samenvatting

114

OH

O

lactobacillic acid

O

O Br

5 mol% L1 5 mol% CuBr.SMe2

2eq C5H11MgBr 75 °C, CH2Cl2, 20 h O

O91%

97% ee

FePPh2

N H

(S,Rfe) Taniaphos L1

PPh2

O

O OOGrubbs II, toluene120 °C, 5 h

81%

Schema 2. De chemische structuur van lactobacilluszuur en centrale stappen van de synthese.

Het tweede deel van dit proefschrift beschrijft de verdere ontwikkeling van een nieuwe methodologie om enanioselectief secundaire en tertiaire alcoholen in één stap te maken. Deze eerder in ons instituut ontwikkelde methodologie is gebaseerd op de koper(I)-gekatalyseerde additie van Grignard reagentia aan aldehyden en ketonen. De ontwikkeling van nieuwe methodologiën is een belangrijk deel van scheikundig onderzoek. Hierdoor worden bestaande processen geoptimaliseerd en nieuwe synthese routes ontwikkeld voor bijvoorbeeld natuurstoffen. Bij het ontwikkelen van een nieuwe methodologie is tegenwoordig niet alleen de regio- en stereoselectiviteit van belang, maar speelt ook de atoomeconomie een rol om de methodologie duurzaam te maken.

In hoofdstuk 4 worden nieuwe resultaten van de koper(I) gekatalyseerde enantioselectieve 1,2-additie van Grignard reagentia aan gesubstitueerde, aromatische aldehyden en aan , -onverzadigde aldehyden beschreven (Schema 3). Eerst zijn de reacties uitgevoerd onder geoptimaliseerde condities die eerder gerapporteerd zijn voor de 1,2-additie aan acetofenonen. Onder deze condities werd een daling van de enantiomere overmaat en de opbrengst genoteerd. Om de enantiomere overmaat en de opbrengst te verhogen zijn verschillende oplosmiddelen en additieven gebruikt.

Samenvatting

115

FePP

H3C H

rev-JosiPhos (S, RFe)-L2

OHO5 mol% CuBr.SMe2

6 mol% L21.3 eq Grignard

MTBE, 78 °C, 14 hH

R Ropbrengst 6-48%

ee 23-67%R: Me, F, CF3

Schema 3. Koper(I)-katalyseerde asymmetrische additive van Grignard reagentia aan aldehyden.

Het werk beschreven in hoofdstuk 5 sluit aan op de eerder gerapporteerde resultaten van de koper(I) gekatalyseerde 1,2-additie van Grignard reagentia aan aromatische ketonen en met name acetofenonderivaten. Voor de ketonen die zijn gebruikt in deze studie werd een significant lagere enantioselectiviteit verkregen. Bovendien werd het overeenkomstige secundaire alcohol als bijproduct gevormd. Het was niet mogelijk de koper(I) gekatalyseerde 1,2-additie op biaryl ketonen toe te passen.

5 mol% CuBr.SMe26 mol% L2

1.3 eq iBuMgBrMTBE, 78 °C, 14 h

R2

O

R2

HO

+ R2

OH

R1 R1 R1

opbrengst 12-50% ee 37-80%

bijproduktR1: F, BrR2: Et, Pr, Bz

Schema 4. Koper(I)-katalyseerde alkylering van aromatischen ketonen met Grignard reagentia.

Samenvatting

116

Zusammenfassung

Zusammenfassung

118

Der erste Teil der Doktorarbeit befasst sich mit der Synthese von Verbindungen, die aus der Zellwand von Bakterien isoliert werden, wie z.B. Lipiden oder Fettsäuren. Die Zellwand grenzt das Innere der Zelle von der Umgebung ab und wirkt somit als Schutzbarriere. Die Zellwand von Tuberkulosebakterien besteht aus sehr speziellen Fettsäuren, wodurch nur ein geringer unkontrollierter Stoffaustausch möglich ist. Die geringe Permeabilität der Zellwand ist ein Grund warum bisherige Therapien noch nicht effizient wirken. Um die genauen Abläufe studieren zu können, benötigen Biochemiker sehr reine Zellwandverbindungen. Die chemische Synthese dieser Verbindungen bietet eine Alternative zur Isolierung aus der Zellwand, da Mycobacterium tuberculosis (der Erreger von Tuberkulose) ein sehr langsam wachsendes Bakterium ist und nur in speziellen Forschungslaboratorien mit entsprechender Zulassung kultiviert werden kann.

Ein weiterer Vorteil der chemischen Synthese ist, dass die einzelnen Zellwandkomponenten gezielt hergestellt werden können. Auf diesem Wege kann man eine hohe Regio- und Stereoselektivität erreichen. Außerdem bietet die chemische Synthese einen zusätzlichen Spielraum zur Derivatisierung der Zellwandkomponenten, die es dem Biochemiker ermöglichen die Prozesse in der Zelle besser zu verfolgen. Zudem ermöglicht die Synthese dieser Naturstoffe eine vollständige Aufklärung der chemische Struktur, durch vergleich mit den isolierten Verbindungen. Diese detaillierten Informationen sind wiederum wichtig, um die zellulären Prozesse auf molekularem Niveau zu verstehen.

Kapitel 2 beschäftigt sich mit der Synthese von -Mykolsäuren (Schema 1), Fettsäuren isoliert aus der Zellwand von Mycobacterium tuberculosis Bakterien. Ein

-Mykolsäure Molekül kann man in drei Bausteine unterteilen: die Mykol Kette, die Meromykol Kette und den -Hydroxycarbonsäure Rest. Nach der retrosynthetischen Analyse war es unser Ziel die zwei Cyclopropanringe in der Meromykol Kette aus dem gleichen Baustein herzustellen. Idealerweise sollte dieser Baustein in einer asymmetrischen, katalytischen Reaktion synthetisiert werden. Das Intermediat sollte zwei verschiedene funktionelle Gruppen besitzen, um dann die verschiedenen Alkylketten einzuführen. Ein möglicher Reaktionstyp, der diese Anforderungen erfüllt, ist die intramolekulare Cyclopropanierung von allylischen Diazoacetaten, beschrieben von Doyle et al.. Leider ergaben sich Probleme beim Ausführen der Reaktion im großen Maßstab. Bei der Weiterführung des Projektes muss daher für die Synthese der Meromykol Kette eine neue Strategie gesucht werden. Die Synthese der -Hydroxycarbonsäure-Funktion wurde anhand von Modelsubstrat (A) untersucht. In einer asymmetrischen Hydrierung wurde -Ketoester A zur

Zusammenfassung

119

entsprechen -Hydroxycarbonsäure B umgesetzt. Die Säure B konnte mit hoher Ausbeute (95%) und mit einer exellenten Enantioselektivität (>95%) isoliert werden.

OH

COOH

mycolic chain

meromycolic chain

O

CHN2

O O O46%87% ee

0.1 mol% Rh2(4S-MEOX)4CH2Cl2, reflux, 12-18 h

Rh2(4S-MEOX)4

Rh Rh

N O

O N COOMe

HMeOOC

H

OH

MeOOC

O N COOMe

H

[(RuCl(T-BINAP))2(μ-Cl)3NH2Me2]EtOH, 20 bar, 50 °C, 25 h

95%>95% ee

O

O O

17O

OH O

17

A B

Schema 1. Chemische Struktur von -Mykolsäure und Schlüsselschritte der Synthese.

Die Synthese einer weiteren Fettsäure „lactobacillic acid“ (Schema 2), zuerst isoliert aus der Zellwand von Lactobacillus arabinosus Bakterien, wird in Kapitel 3 beschrieben. Diese Fettsäure besitzt wie Mykolsäuren einen Cyclopropanring als funktionelle Gruppe. Bei dieser Synthese wurde eine andere Strategie verfolgt. Der cis-konfigurierte Cyclopropanring soll in einer Simmons-Smith-Reaktion entstehen. Der chirale Allylalkohol, das Startmaterial für die Simmons-Smith Reaktion, wurde in mehreren Schritten hergestellt. In einer hetero asymmetische allylische Alkylierung wurde der chirale Alkohol synthetisiert. Die cis-konfigurierte Doppelbindung entstand in einer Ringschlussmetathese eines Makrolaktones. Durch eine Präkoordination des Zinkcarbenoids mit dem chiralen Alkohol konnte die Stereoinformation auf den Cyclopropanring übertragen werden. Letzten Endes konnte die Synthese von „lactobacillic acid“ nicht beendet werden, da alle Versuche die Hydroxylgruppe zu eliminieren scheiterten.

Zusammenfassung

120

OH

O

lactobacillic acid

O

O Br

5 mol% L1 5 mol% CuBr.SMe2

2eq C5H11MgBr 75 °C, CH2Cl2, 20 h O

O91%

97% ee

FePPh2

N H

(S,Rfe) Taniaphos L1

PPh2

O

O OOGrubbs II, toluene120 °C, 5 h

81%

Schema 2. Chemische Struktur von „lactobacillic acid“ und Schlüsselschritte der Synthese.

Ein zweiter Themenbereich dieser Doktorarbeit ist die Weiterentwicklung einer neuen Methodologie zur Synthese chiraler sekundäre und tertiäre Alkohole. Dazu verwendet man bei der neuen Methodologie die Kupfer(I)-katalysierte asymmetrische Addition von Grignardreagenzien zu Aldehyden oder Ketonen. Die Entwicklung neuer Methoden ist ein wesentlicher Teil der chemischen Forschung, um bestehende Vorschriften zu optimieren oder neue Synthesewege für die Synthese von Naturstoffen zu finden. Heutzutage spielt bei der Methodenentwicklung nicht nur die Regio- und Stereoselektivität eine Rolle, sondern es muss auch die Atomökonomie für die Nachhaltigkeit des Prozesses berücksichtigt werden.

Kapitel 4 handelt von der Kupfer(I)-katalysierten Addition von Grignardreagenzien zu aromatischen Aldehyden und , -ungesättigten Aldehyden (Schema 3). Zu Beginn des Projektes wurden die Versuche unter den Standardversuchsbedingungen durchgeführt, die zuvor in der 1,2-Addition von Grignardreagenzien zu Acetophenonen optimiert wurden. Jedoch wurde unter diesen Bedingungen für Aldehyde eine deutlich geringere Ausbeute (6-48%) und eine wesentlich schlechtere Enantioselektivität (37-80%) beobachtet. Ein Grund dafür ist die schnelle Hintergrundreaktion. Daher wurden verschiedene Lösungsmittel und Additive getestet, um die Hintergrundreaktion zu unterdrücken und die Enantioselektivität zu erhöhen.

Zusammenfassung

121

FePP

H3C H

rev-JosiPhos (S, RFe)-L2

OHO5 mol% CuBr.SMe2

6 mol% L21.3 eq Grignard

MTBE, 78 °C, 14 hH

R RAusbeute 6-48%

ee 23-67%R: Me, F, CF3

Schema 3. Kupfer(I)-katalysierte Alkylierung von Aldehyden mit Grignardreagenzien.

In Kapitel 5 wird dieselbe Methodologie auf aromatische Ketone mit längeren Alkylketten angewendet. Im Gegensatz zu den bisher verwendeten Acetophenonderivate, zeigen die in dieser Studie verwendeten Ketone eine deutlich geringere Enantioselektivität (Schema 4). Zusätzlich wird die Entstehung eines Nebenproduktes, die Bildung des sekundären Alkohols, beobachtet. Die Grenze der Anwendungsbreite der Kupfer(I)-katalysierten Alkylierung mit Grignardreagenzien wurde erreicht, als Biarylketone als Startmaterial verwendet wurden.

5 mol% CuBr.SMe26 mol% L2

1.3 eq iBuMgBrMTBE, 78 °C, 14 h

R2

O

R2

HO

+ R2

OH

R1 R1 R1

Ausbeute 12-50% ee 37-80%

NebenproduktR1: F, BrR2: Et, Pr, Bz

Schema 4. Kupfer(I)-katalysierte Addition von Ketonen mit Grignardreagenzien.

Zusammenfassung

122

Acknowledgements

Acknowledgements

124

After my arrival in Groningen I realized quite fast, how special this place is. And I mean not only Groningen as a city with its lively, colorful market, nice cafés and many festivals, but also the Stratingh institute. For me this is one of the best places to perform research, because of the excellent equipment, the collaborations between the research groups and the particular working atmosphere. This atmosphere is created by many great individuals forming this dynamic group of international researchers from around the world. Therefore, I would like to thank everybody, who was part of this group and made my five years of PhD to a unique experience.

Thank you Adri, for giving me the opportunity to work in your group, and to be always supportive. I am grateful, that you gave me the chance to work on the school campaign promoting the chemistry department at German high schools. I thank you for taking the time, to discuss with me from the first moment everything in Dutch. I appreciate very much, your time for the corrections and especially the last minute corrections!

I would like to acknowledge the members of the reading committee, Prof. Harutyunyan, Prof. Heeres and Prof. Pieters for reading and approving my thesis as well as for their valuable corrections of the manuscript.

I would like to express my gratitude to all technicians for their help: Ebe for teaching me how to use the hydrogenation system and for taking care of all our equipment, Theodora and Monique for teaching me all the details about HPLC, GC-MS and measuring my HRMS samples, Hans for measuring elemental analysis, Pieter and Wim for all problems related to the NMR, especially in the new building. Thank you, Hilda and Alphons for your advice with documents and paperwork.

The first two and half years of my PhD I spend on Koffie Klub Island, lab 14.245 in Nijenborgh 4. I shared the lab with Adi (Adipedia), Bart, Danny (Singstar), Diane, Edwin, Fabian, Jeffrey, Leticia, Maxime, Niek, Santi (la Chouffe), Wendy and Yange. I thank you all for making my everyday life in the lab so enjoyable. I especially thank our neighboring lab (Vysom) for providing the famous Friday afternoon music mix and sharing chemicals: Ashoka, Bas, Cati, Chris, Dorus, Felix, Johannes, Manuel, Maria, Miro, Thomas and Simon. In 2011 Adri’s group moved to the green building, Linnaeusborg, together with Anna’s group and some people from Ben’s group. I want thank everyone for creating a nice working atmosphere in the new labs: Alrik, Anja, Ana, Anniek, Bea, Blijke, Claudia, Derk Jan, Gosha, Hylke, Jasmin, Jelle, Jonas, Kiran, Marcel, Mark, Mickel, Milon, Nick, Patrick, Peter, Selma, Stephan, Steven, Tiziana, Vasu, Wienand, Wiktor, Wim and Zhongtao.

Acknowledgements

125

During my PhD more research groups joined the Stratingh institute, therefore, I will not name anyone in person. I want to thank everyone, who organized or participated in one or the other social activity (work week, bbq’s, rondje lab, sailing trip, christmas borrel, soccer subgroup), for creating those unforgettable moments.

I want to thank Hella, Anne and Jeffrey Bos for their patience to chat with me in Dutch, especially in the beginning. This really encouraged me to continue speaking Dutch. Later in the Linnaeusborg I was lucky to share the office with Edwin, Anniek and Mickel, who tried to teach me small details and Dutch sayings. Thank you!

Katja, Maria, and Zhongtao I am grateful for your time to proof read my manuscript. Pat, Felix and Rik thank you for correcting my summary and for your useful suggestions. Anne, I appreciate your help with rewriting the Dutch summary!

Thank you, Erik, Cati and Jochem for your pleasant welcome and support, when I started in Syncom.

Also outside of the lab I really enjoyed my life, during many nice dinners, board game evenings, concerts, festivals, pilates, dance shows, trips, cooking classes with Suresh, parties, and the traditional Saturday coffee meetings. I am very glad to have shared those moments with good friends!

I am happy two have two of my best friends as paranimfs at my side on a very special day. Pat, thank you for all your support during the last couple of years, and for organizing our wonderful trip to Thailand. Tizi, thank you for sharing so many stories form the past and the present, and for creating the expression German vibrations, it always makes me smile

Liebe Oma, leider konnte ich in den letzten Monaten nicht so häufig zu Hause sein, um dich und Mama besser zu unterstützen. Ich bewundere deinen Kampfgeist. Danke für deine vielen Anrufe und Besuche hier in Groningen!

Liebeste Mama, ich danke dir für all die Liebe und Unterstützung, die du mir gegeben hast, ohne die hätte ich diese Arbeit nie zu Ende geschrieben!

Carpe diem