Modern Methods in Stereoselective Aldol Reactions E-Book
156,99 €
Mehr erfahren.
- Herausgeber: Wiley-VCH
- Kategorie: Wissenschaft und neue Technologien
- Sprache: Englisch
The selective formation of bondings between molecules is one of the major challenges in organic chemistry, and the so-called aldol reaction is one of the most important for this purpose. These reactions are a highly useful tool for developing such novel substances as natural products and pharmaceuticals.
Likes its highly successful and much appreciated predecessor, "Modern Aldol Reactions", this ready reference provides a systematic overview of methodologies for installing a required configuration during an aldol addition step, but shifts the focus so as to cover the latest developments.
As such, it presents a set of brand new tools, including vinylogous Mukaiyama-aldol reactions and substrate-controlled aldol reactions, as well as asymmetric induction in aldol additions. Furthermore, new developments in existing stereoselective aldol additions are described, such as the deployment of supersilyl groups or organocatalyzed aldol additions. All of these methodologies are presented in the context of their deployment in the total synthesis of natural products.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 724
Veröffentlichungsjahr: 2013
Ähnliche
Table of Contents
Related Titles
Title Page
Copyright
Preface
List of Contributors
Chapter 1: Stereoselective Acetate Aldol Reactions
1.1 Introduction
1.2 Mukaiyama Aldol Reaction
1.3 Metal Enolates
1.4 Conclusions
References
Chapter 2: The Vinylogous Mukaiyama Aldol Reaction in Natural Product Synthesis
2.1 Introduction
2.2 Aldehyde-Derived Silyl Dienol Ethers
2.3 Ester-Derived Silyl Dienol Ethers
2.4 Amide-Derived Silyl Dienol Ethers – Vinylketene Silyl N,O-Acetals
2.5 Acyclic Acetoacetate-Derived Silyl Dienolates – Chan's Diene
2.6 Cyclic Acetoacetate-Derived Dienolates
2.7 Furan-Derived Silyloxy Dienes
2.8 Pyrrole-Based 2-Silyloxy Dienes
2.9 Comparison with Other Methods
References
Chapter 3: Organocatalyzed Aldol Reactions
3.1 Introduction
3.2 Proline as Organocatalyst
3.3 Proline Derivatives as Organocatalysts
3.4 Conclusions and Outlook
References
Chapter 4: Supersilyl Protective Groups in Aldol Reactions
4.1 Introduction
4.2 Aldol Addition with Acetaldehyde-Derived Super Silyl Enol Ether (1)
4.3 α-Substituted Silyl Enol Ethers Derived from Aldehydes
4.4 Aldol Addition to Chiral Aldehydes
4.5 One-Pot Sequential Aldol Reactions
4.6 Sequential Aldol–Aldol Reactions of Acetaldehyde
4.7 Double Aldol Reactions with α-Substituted Silyl Enol Ethers
4.8 Stereochemical Considerations
4.9 Aldol Reactions of β-Supersiloxy Methyl Ketones
4.10 Total Synthesis of Natural Products Using Supersilyl Aldol Reactions
4.11 Conclusion and Outlook
References
Chapter 5: Asymmetric Induction in Aldol Additions
5.1 Introduction
5.2 Asymmetric Induction Using Chiral Ketones
5.3 Asymmetric Induction Using Chiral Aldehydes
5.4 Asymmetric Induction in the Aldol Addition of Chiral Enolates to Chiral Aldehydes
References
Chapter 6: Polypropionate Synthesis Via Substrate-Controlled Stereoselective Aldol Couplings of Chiral Fragments
6.1 Introduction
6.2 Principles of Stereoselective Aldol Reactions
6.3 Stereoselective Aldol Coupling of Chiral Reactants
6.4 2-Alkoxyalkyl Ethyl Ketones: 2-Desmethyl Polypropionate Equivalents
6.5 Conclusions
References
Chapter 7: Application of Oxazolidinethiones and Thiazolidinethiones in Aldol Additions
7.1 Introduction
7.2 Preparation of Oxazolidinethione and Thiazolidinethione Chiral Auxiliaries
7.3 Acylation of Oxazolidinethione and Thiazolidinethione Chiral Auxiliaries
7.4 Propionate Aldol Additions
7.5 Acetate Aldol Additions
7.6 Glycolate Aldol Additions
References
Chapter 8: Enzyme-Catalyzed Aldol Additions
8.1 Introduction
8.2 Pyruvate Aldolases
8.3 N-Acetylneuraminic Acid Aldolase (NeuA)
8.4 Dihydroxyacetone Phosphate (DHAP) Aldolases
8.5 D-Fructose-6-Phosphate Aldolase and Transaldolase B Phe178Tyr: FSA-Like Aldolases
8.6 2-Deoxy-D-Ribose-5-Phosphate Aldolase (RibA or DERA; EC 4.1.2.4)
8.7 Glycine/Alanine Aldolases
8.8 Aldol Reactions Catalyzed by Non aldolases
8.9 Conclusions and Perspectives
References
Index
Related Titles
Majumdar, K. C., Chattopadhyay, S. K. (eds.)
Heterocycles in Natural Product Synthesis
2011
ISBN: 978-3-527-32706-5
Poupon, E., Nay, B. (eds.)
Biomimetic Organic Synthesis
2011
ISBN: 978-3-527-32580-1
Nicolaou, K. C., Chen, J. S.
Classics in Total Synthesis III
Further Targets, Strategies, Methods
2011
ISBN: 978-3-527-32958-8
Yang, J
Six-Membered Transition States in Organic Synthesis
2010
ISBN: 978-0-470-65258-9
Joule, J. A., Mills, K.
Heterocyclic Chemistry
2010
ISBN: 978-0-470-65258-9
Zabicky, J. (ed.)
The Chemistry of Metal Enolates
2009
ISBN: 978-0-470-06168-8
Carreira, E. M., Kvaerno, L.
Classics in Stereoselective Synthesis
2009
ISBN: 978-3-527-32452-1
The Editor
Prof. Dr. Rainer Mahrwald
Humboldt-Universität Berlin
Institut für Chemie
Brook-Taylor-Str. 2
12489
Berlin
All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.
Library of Congress Card No.: applied for
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library.
Bibliographic information published by the Deutsche Nationalbibliothek
The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at <http://dnb.d-nb.de>.
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Boschstr. 12, 69469 Weinheim, Germany
All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form — by photoprinting, microfilm, or any other means — nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law.
Print ISBN: 978-3-527-33205-2
ePDF ISBN: 978-3-527-65674-5
ePub ISBN: 978-3-527-65673-8
oBook ISBN: 978-3-527-65671-4
mobi ISBN: 978-3-527-65672-1
Cover Design Adam-Design, Weinheima
Typesetting Laserwords Private Limited, Chennai, India
Preface
Stereoselectivity is one of the most important aspects for natural product chemists. Following the increasing possibility of detection and assignment of stereogenic centers, a tremendous increase in stereoselective methods of organic reactions, particularly aldol reactions, has been noticed. In the beginning of this development, only sporadic examples of stereoselective aldol reactions were described, mostly in the context of total syntheses of natural products. An outstanding early example is the R. B. Woodward's proline-catalyzed aldol addition in the total synthesis of erythronolide A at the Harvard University in 1981. In the following three decades, a vast arsenal of stereoselective aldol additions has been developed (see Figure).
This book provides a comprehensive review of modern aldol reactions, especially in the aspect of how to achieve high stereoselectivity – diastereoselectivity as well as enantioselectivity. Stereoselection is discussed under several different aspects. One aspect is the deployment of different substrates – acetate or propionate aldol reactions. Another aspect is the mode of action including metal enolate chemistry, Lewis acid as well as Lewis base catalysis, enzymatic catalysis, and organocatalysis. There are some overlappings of these aspects in the chapters covering the cross-cutting themes of vinyloguos Mukaiyama reaction or asymmetric inductions (e.g., compare Scheme 1.50 with Scheme 2.59) or total synthesis of dolastatin 19 – (compare Scheme 1.82 with Scheme 5.8). These overlappings, however, are intentional in order to give a comprehensive insight into the techniques for installing required configurations during aldol reactions. The utility of the corresponding methods is shown in the context of total syntheses of natural products. All chapters are thoroughly well written by experts in the respective fields.
It is my pleasure to express profound gratitude to the 15 authors for their huge endeavor to organize and summarize this vast amount of material. It has been a great pleasure for me to work with this team of authors at all times. Finally, my special thanks go to Elke Maase and Bernadette Gmeiner at WILEY for their fine work in making this book a reality.
Berlin, Autumn 2012
Rainer Mahrwald
List of Contributors
Chapter 1
Stereoselective Acetate Aldol Reactions
Pedro Romea and Fèlix Urpí
1.1 Introduction
The stereochemical control of aldol reactions from unsubstituted enol- or enolatelike species, what are known as acetate aldol reactions, has been a matter of concern for nearly 30 years [1, 2]. Indeed, pioneering studies soon recognized that the asymmetric installation of a single stereocenter in such aldol reactions was much more demanding than the simultaneous construction of two new stereocenters in the related propionate counterparts (Scheme 1.1) [3]. This challenge, together with the ubiquitous presence of chiral β-hydroxy α-unsubstituted oxygenated structures in natural products, has motivated the development of new concepts and strategies and a large number of highly stereoselective methodologies. These involve Lewis-acid-mediated additions of enolsilane derivatives of carbonyl compounds to aldehydes (Mukaiyama aldol variant) [4, 5], a plethora of transformations that take advantage of the reactivity of boron, titanium(IV), and tin(II) enolates (metal enolates) [6], and some insightful organocatalytic approaches [7]. In spite of these accomplishments, the quest for more powerful and selective methodologies and a better understanding of their intricate mechanisms is an active area of research. Herein, we describe the most significant achievements in the field of stereoselective acetate aldol reactions based on the Lewis-acid-mediated addition of enolsilanes and metal enolates to aldehydes, with particular attention to their application to the asymmetric synthesis of natural products. Recent advances in parallel organocatalytic procedures are not discussed.
Scheme 1.1 Aldol reactions.
1.2 Mukaiyama Aldol Reaction
1.2.1 Concept and Mechanism
With some significant exceptions, enolsilanes are unreactive toward aldehydes.1 This lack of reactivity can be overcome by increasing the electrophilic character of aldehydes or the nucleophilicity of enolsilanes. The former option is achieved by coordination of Lewis acids (MLn) to the carbonyl group, which enhances the electrophilicity of the C—O bond and facilitates the attack of enolsilanes. This represents the canonical Mukaiyama aldol variant ((1) in Scheme 1.2) [4, 5]. It also covers vinylogous aldol transformations, which involve the reactions of γ-unsubstituted β, γ-conjugated enolsilanes ((2) in Scheme 1.2) [8]. In turn, the latter option takes advantage of the activation of the nucleophilic character of enolsilanes by binding of Lewis bases such as phosphoramides (O—P(NR2)3) to the silicon atom ((3) in Scheme 1.2) [9].
Scheme 1.2 Mukaiyama aldol variants.
Early mechanistic analyses suggested that Lewis-acid-mediated aldol reactions represented in Scheme 1.2 proceeded through open transition states [4, 5, 10]. This model assumes a transoid geometry for the Lewis-acid-aldehyde complex, which the enolsilane attacks following antiperiplanar or synclinal approaches, as represented in Scheme 1.3. Antiperiplanar transition states I and II are usually more favorable because of the minimization of dipolar interactions, the steric interactions between the enolsilane (R1 or R3SiO groups) and the aldehyde (R2 group) being the main source of instability. Similar steric interactions arise in synclinal transition states III and IV, whereas V and VI are characterized by a destabilizing interaction between the enolsilane and the Lewis acid coordinated to the carbonyl oxygen. Then, steric and stereoelectronic interactions determine the relative stability of I–VI and the capacity to differentiate one from the other faces of the carbonyl bond.
Scheme 1.3 Open transition states for Mukaiyama aldol reactions.
Despite the importance and utility of this paradigm, it is probably an oversimplified model because it ignores the fate of the silyl group. In this respect, some models take into account the silicon moiety and suggest cyclic transition states VII–IX, as represented in Scheme 1.4. Importantly, the role of the silyl group is not limited to influencing the nature of the transition state, because the silicon transfer from the enolsilane to the β-alkoxy position may be a key step in the overall mechanism and becomes crucial to the turnover necessary for nonstoichiometric transformations [11].
Scheme 1.4 Cyclic transition states for Mukaiyama aldol reactions.
Irrespective of the mechanistic pathway, the asymmetric induction achieved by these Lewis-acid-mediated aldol reactions depends on chiral elements on the enolsilane (the nucleophilic partner), the aldehyde (the electrophilic partner), or the Lewis acid (the activating element), so they must all cooperate to provide the appropriate face differentiation of the carbonyl bond in order to control the configuration of the new stereocenter. The influence of these elements is discussed in the following sections.
1.2.2 Chiral Auxiliaries
In the context of emergence of chiral auxiliaries as powerful platforms to achieve asymmetric transformations, Helmchen reported highly diastereoselective aldol reactions of chiral auxiliary-based silyl ketene acetals (1) and (2) [12, 13]. As shown in Scheme 1.5, TiCl4-mediated additions of 1 and 2 to isobutyraldehyde afforded aldol adducts 3 and 4 in good yields and excellent diastereomeric ratios, presumably through a chairlike transition state in which the titanium atom is simultaneously coordinated to the carbonyl and the OTBS group. In turn, these adducts can be converted into the corresponding β-hydroxy acids in quantitative yield by simple treatment with KOH in methanol.
Scheme 1.5 Chiral auxiliary-based Mukaiyama aldol reactions.
This approach was quickly surpassed by alternative methodologies based on chiral aldehydes or Lewis acids and bases (Sections 1.2.4–1.2.6). Nevertheless, new findings restored the interest in this sort of transformations a few years ago. Indeed, Kobayashi described highly stereoselective vinylogous Mukaiyama aldol reactions using silyl vinyl ketene N,O-acetals prepared from valine-derived 1,3-oxazolidin-2-ones [14]. As represented in Scheme 1.6, TiCl4-mediated additions of α-methyl acetal (5) to aliphatic, α, β-unsaturated, and aromatic aldehydes afforded δ-hydroxy-α-methyl-α, β-unsaturated imides (6) in excellent yields and diastereomeric ratios. Such outstanding remote asymmetric induction was believed to arise from a conformation in which the chiral heterocycle is almost perpendicular to the dienol plane and the isopropyl group overhangs the upper face of the dienol moiety. Then, the aldehyde approaches from the less hindered face through an open transition state in which the α-methyl group appears to be essential to achieve the observed high stereocontrol. Finally, the chiral auxiliary can be removed by well-known methodologies used for Evans auxiliaries.
Scheme 1.6 Chiral auxiliary-based vinylogous Mukaiyama aldol reactions.
This methodology was used for the construction of the AB ring of fomitellic acids (7) (Scheme 1.7) [15]. Initially, application of the standard conditions to ent-5 and enal (8) provided the desired aldol (9), but the reaction was slow and hard to reproduce. Then, a thorough study of this particular reaction uncovered significant rate enhancements by adding catalytic amounts of water in toluene, which permitted to obtain aldol (9) in 76% yield as a single diastereomer in a straightforward and consistent way [16]. The origin of this catalytic effect remains unclear, but it has proved to be general and has been successfully applied to other aldehydes [17].
Scheme 1.7 Synthesis of the central ring of fomitellic acids.
1.2.3 Chiral Methyl Ketones
There are no systematic studies on the asymmetric induction imparted by chiral methyl ketones. However, most of the examples reported so far suggest that substrate-controlled Mukaiyama aldol reactions based on chiral methyl ketones are poorly stereoselective. This lack of stereocontrol is well illustrated by the aldol reaction of chiral silyl enol ether (10), in which the major diastereomer, 11 or 12, depends on the achiral aldehyde (Scheme 1.8) [18].
Scheme 1.8 Asymmetric induction imparted by a silyl enol ether from a chiral methyl ketone.
In a more complex framework, De Brabander also reported that silyl enol ethers from enantiomeric methyl ketones (13) and (ent-13) underwent additions to chiral aldehyde (14) to afford the corresponding aldol adducts 15 and 16 in similar yields (Scheme 1.9) [19]. Considering that the new C11-stereocenters possess the same configuration in both adducts and that the diastereoselectivity is comparable for both processes, it can be concluded that the asymmetric induction provided by the aldehyde is much more important than that provided by the ketone.
Scheme 1.9 Chiral methyl ketones in stereoselective Mukaiyama aldol reactions.
Lactate-derived and other α-hydroxy methyl ketones are exceptions to this trend. Thus, Trost found that TiCl4-mediated aldol reactions of pivaloyl-protected silyl enol ether (17) afforded β-hydroxy ketones (18) in high yields and diastereomeric ratios up to 98 : 2 [20]. This was assumed to be achieved through an eight-membered cyclic transition state in which the titanium is simultaneously bound to the aldehyde and the enolsilane ((1) in Scheme 1.10). Importantly, dipole–dipole interactions are understood to favor the antiperiplanar arrangement of the C–OPiv and the C–OSi bonds and impel the aldehyde toward the less hindered face of the enolsilane. Moreover, Kalesse reported that parallel BF3-catalyzed additions of silyl enol ethers (19) to isobutyraldehyde afforded the corresponding aldols (20) with excellent diastereoselectivity but in low yield ((2) in Scheme 1.10) [21]. The origin of such remarkable stereocontrol is unclear.
Scheme 1.10 Asymmetric induction imparted by chiral α-hydroxy methyl ketones in Mukaiyama aldol reactions.
At this point, it is worth mentioning that Yamamoto has also reported highly diastereoselective Mukaiyama aldol reactions based on chiral β-tris(trimethylsilyl)silyloxy methyl ketones containing a single stereocenter at the β-position [22]. This chemistry is discussed in connection with parallel methodologies (Sections 1.2.4 and 1.3.6).
Finally, Ley reported that reactions of silyl enol ethers from chiral π-allyltricarbonyliron lactone or lactam complexes proceeded with a significant remote stereocontrol [23]. This is illustrated by the BF3-mediated addition of complexes 21 to benzaldehyde furnishing β-hydroxy ketones (22) with excellent diastereomeric ratios (Scheme 1.11). The remarkable 1,7-induction provided by these substrates is due to the chiral environment created on the lower face of the silyl enol ether (the upper face is blocked by the tricarbonyliron moiety) by the endo-oriented methyl substituent at the sp3-stereocenter. Then, the incoming activated aldehyde approaches in a synclinal arrangement in which unfavorable steric interactions are minimized.
Scheme 1.11 Asymmetric induction imparted by chiral π-allyltricarbonyl iron complexes in Mukaiyama aldol reactions.
As the iron lactone and lactam (22) can be easily decomplexed to afford a rich array of stereodefined derivatives, this reaction may represent a powerful tool to the rapid construction of highly functionalized systems under remote stereocontrol. For instance, total synthesis of (−)-gloeosporone (23) commenced with the addition of silyl enol ether from methyl ketone (24) to benzyloxypropanal, which afforded aldol (25) as a single diastereomer in a 63% yield (Scheme 1.12) [24].
Scheme 1.12 Synthesis of (−)-gloeosporone.
1.2.4 Chiral Aldehydes
The asymmetric induction of chiral aldehydes in Mukaiyama aldol reactions is much more important and has stimulated the formulation of increasingly more refined models to predict the π-facial selectivity in nucleophilic additions to the carbonyl bond [25]. Therefore, the influence of α- and β-substituents has received particular attention and is described in detail in the following sections.
1.2.4.1 1,2-Asymmetric Induction
Pioneering studies on acyclic stereoselection established that Mukaiyama acetate aldol additions of enolsilane derivatives (26) and (27) to chiral α-methyl aldehydes (28) proceeded with high diastereofacial selectivity to favor 3,4-syn aldol adducts (29) (Scheme 1.13) [26].
Scheme 1.13 Asymmetric induction imparted by chiral α-methyl aldehydes in Mukaiyama aldol reactions.
As expected, the 1,2-asymmetric induction of such aldehydes was eroded when R2 was sterically similar to the α-methyl. The challenge posed by these transformations can be met by using more bulky nucleophiles, as has been observed in the aldol additions of enolsilanes (26), (27), and (30) to 2-methyl-3-phenylpropanal (Scheme 1.14). The stereochemical outcome of these reactions shows that enhancement of the steric hindrance of R1 and SiR3 groups gives the corresponding 3,4-syn aldols (31) in higher diastereomeric ratios [26, 27]. A parallel improvement can also be attained by employing more bulky Lewis acids, but steric influences must be analyzed carefully because some combinations of bulky nucleophiles and Lewis acids do not provide the expected results [27].
Scheme 1.14 Mukaiyama aldol additions to (R) 2-methyl-3-phenylpropanal.
The Felkin–Anh model [25, 28] is usually invoked to account for the asymmetric induction observed in the Mukaiyama aldol additions to these chiral α-methyl aldehydes. Thus, once the methyl group has been identified as the medium size group, the major 3,4-syn diastereomer is obtained by bringing the enolsilane close to the face of the C—O bond in which the steric interactions between the nucleophile and the α-substituent (H vs Me) are weaker (X in Scheme 1.15).
Scheme 1.15 The Felkin–Anh model for Mukaiyama aldol additions to chiral α-methyl aldehydes.
The total synthesis of borrelidin (32) reported by Theodorakis contains a good example of stereocontrol based on the asymmetric induction imparted by such chiral aldehydes [29]. As represented in Scheme 1.16, the Mukaiyama aldol addition of silyl ketene acetal (27c) to α-methyl aldehyde (33) produced the desired ester (34) as a 80 : 20 mixture of diastereomers in a 95% combined yield, which demonstrates the π-facial selectivity provided by the α-stereocenter of aldehydes of this kind [30].
Scheme 1.16 Synthesis of borrelidin.
The substitution of the methyl group by a heteroatom affects these transformations dramatically. Indeed, a tenet in asymmetric synthesis states that nucleophilic additions to chiral aldehydes bearing an α-heteroatom attain outstanding levels of stereocontrol provided that the reaction is carried out under conditions in which chelate organization is favored. In this context, the Cram model [25, 31] accounts for the stereochemical outcome of chelate-controlled Mukaiyama aldol reactions. According to this model, the appropriate choice of the Lewis acid and the protecting group of α-hydroxy aldehydes permits the formation of stable five-membered chelated complexes and gives the corresponding 3,4-syn aldol adducts in a highly diastereoselective manner, presumably through an open transition state in which the nucleophile approaches the less hindered face of the chelated carbonyl group (Scheme 1.17).
Scheme 1.17Cram model for Mukaiyama aldol additions to chiral α-hydroxy aldehydes.
This highly reliable and powerful element of stereocontrol has been widely exploited in the synthesis of natural products. For instance, Sunazuka and Omura used a chelate-controlled Mukaiyama aldol reaction for the total synthesis of an epimer of guadinomine C2 (35). As shown in Scheme 1.18, addition of silyl ketene S,O-acetal (36a) to chiral α-OPMB aldehyde (37) in the presence of 1.1 equiv of TiCl3(i-PrO) gave 3,4-syn aldol (38) with exceptional diastereoselectivity in 64% yield [32].
Scheme 1.18 Synthesis of (3′S) epimer of guadinomine C2.
Moreover, Forsyth reported that the addition of mild Lewis acid MgBr2 · OEt2 to a mixture of chiral silyl enol ether (39) and α-OPMB aldehyde (40) triggered a smooth aldol reaction that furnished 3,4-syn aldol (41) as a single diastereomer in 79% yield, which was further elaborated to a hapten for azaspiracids (Scheme 1.19) [33]. A very similar transformation was also reported by Evans [34, 35].
Scheme 1.19 Synthesis of a hapten for azaspiracids.
In the absence of chelated intermediates, nucleophilic additions to chiral aldehydes possessing an α-heteroatom are currently explained by the polar Felkin–Anh [36] and Cornforth models [37], which apply to conformations XII–XV arising from rotation about the C1–C2 bond of the aldehyde (Scheme 1.20) [25]. The polar Felkin–Anh model is based on the premise that staggered transition states positioning the C–X bond perpendicularly to the carbonyl bond are preferred ((1) in Scheme 1.20). In turn, the Cornforth model embraces the assumption that electrostatic effects are instrumental in dictating a nearly antiparallel relationship between the carbonyl and the C–X bond ((2) in Scheme 1.20). Then, the stereochemical outcome of these additions depends on the steric interactions between the nucleophile and the remaining α-substituents in alternative transition states. Application of both models to Mukaiyama aldol reactions predicts the preferential formation of 3,4-anti aldol adducts.
Scheme 1.20 Models for Mukaiyama aldol additions to chiral aldehydes possessing an α-heteroatom.
Most of the aldol reactions involving such aldehydes proceed in accordance with these expectations, but systematic studies on the addition of enolsilanes (42) to α-chloro, α-hydroxy, and α-amino aldehydes (43–45) revealed that their diastereoselectivity is dependent significantly on the α-heteroatom and the steric bulk of nucleophiles (Scheme 1.21) [38–40]. Thus, additions of acetone-derived enolsilane (42a) to aldehydes (43) and (44) possessing an electronegative α-heteroatom such as chlorine or oxygen afforded the corresponding 3,4-anti aldols (46a) and (47a) (dr 60 : 40 and 82 : 18, respectively), whereas more sterically hindered pinacolone-derived enolsilane (42c) gave under the same conditions 3,4-syn aldol (46c) or equimolar mixtures of 3,4-anti and syn diastereomers (47c). In turn, N-benzyl-N-tosyl protected valinal (45) always furnished 3,4-anti aldols (48) in modest to excellent diastereomeric ratios (Scheme 1.21).
Scheme 1.21 Mukaiyama aldol additions to chiral aldehydes possessing an α-heteroatom.
In view of these and a few related studies on the influence of bulky Lewis acids, Somfai proposed that N-protected valinal (45) prefers the polar Felkin–Anh manifold, which is not seriously affected by the size of the nucleophile ((1) in Scheme 1.20) [39]. In turn, Mukaiyama aldol additions to α-chloro aldehyde (43) would proceed essentially through antiperiplanar Cornforth transition state XVI provided that R1 is small enough to avoid serious steric interactions with chlorine (Scheme 1.22). Facing such deleterious interactions, sterically hindered silyl enol ethers would react through antiperiplanar anti-Cornforth transition state XVII. A similar rationale could be applied to α-silyloxy aldehyde (44).
Scheme 1.22Cornforth model for Mukaiyama aldol additions to (S) 2-chloro-3-methylbutanal.
1.2.4.2 1,3-Asymmetric Induction
Mukaiyama aldol additions to chiral α-unsubstituted aldehydes bearing a β-heteroatom usually proceed with significant levels of 1,3-anti asymmetric induction [41]. As illustrated in Scheme 1.23, most of the aldol reactions of silyl enol ethers (42) with β-hydroxy, β-chloro, and β-azido aldehydes (49) produce the corresponding 3,5-anti aldols (50) in high diastereomeric ratios regardless of the chelating ability of the Lewis acid and the hydroxyl protecting group [42–44].
Scheme 1.23 Asymmetric induction imparted by chiral α-unsubstituted aldehydes bearing a β-heteroatom in Mukaiyama aldol reactions.
The stereochemical outcome of these reactions may not be readily interpreted because both open-chain and chelation control lead to the same 3,5-anti diastereomer (50). Thus, Evans proposed, after a systematic and insightful study, a revised induction polar model and a chelate-controlled model to account for the high 1,3-asymmetric induction provided by such chiral aldehydes [43]. The former is an open-chain model ((1) in Scheme 1.24) derived from several assumptions common to the Felkin–Anh analysis for 1,2-asymmetric induction. First, it is assumed that torsional effects dictate that aldehyde transition state conformations adopt a staggered relationship between the forming bond and the aldehyde α-substituents. Second, it is also assumed that the principal diastereomer arises from the reactant-like transition state wherein the β-stereocenter is oriented anti, rather than gauche, to the forming bond because this geometry reduces nonbonded interactions between the aldehyde α-substituents and the incoming nucleophile. In turn, the chelate-controlled model applies to suitable β-hydroxy-protected aldehydes able to form stable six-membered chelates, which can subsequently react through transition states involving either boat or half-chair conformations ((2) in Scheme 1.24).
Scheme 1.24Evans models for Mukaiyama aldol additions to chiral α-unsubstituted aldehydes bearing a β-heteroatom.
Regardless of mechanistic considerations, the predictable and high stereocontrol provided by these aldehydes has been used in many total syntheses. Evans reported in the total synthesis of bryostatin 2 (51) that aldol coupling of bis(trimethylsilyl)dienol ether (52a) with β-OPMB aldehyde (53) ((1) in Scheme 1.25) was only modestly stereoselective when it was carried out in the presence of MgBr2 · OEt2 or BF3 · OEt2, whereas strong Lewis acid (TiCl4, SnCl4) did not effect a clean transformation. Alternatively, mild TiCl2(i-PrO)2 in CH2Cl2 at − 78 °C delivered aldol (54) in a high-yielding and stereoselective manner (dr 86 : 14, 93%). Importantly, the diastereoselectivity was improved by using toluene (dr 94 : 6, 83%), a result that is consistent with the operation of electrostatic effects as the stereochemical control element [45]. In turn, Panek took advantage of the 1,3-asymmetric induction of β-alkoxy aldehydes in a challenging Mukaiyama aldol reaction to assemble an advanced intermediate in the total synthesis of leucascandrolide A (55) [46]. Indeed, the coupling of chiral silyl enol ether (56) and aldehyde (57) ((2) in Scheme 1.25) afforded the desired anti aldol (58) in 81% yield and excellent diastereomeric ratio (dr > 94 : 6). This, presumably occurred through an open transition state in which the β-stereocenter of the aldehyde determines the approach of the nucleophile to the carbonyl in accordance with the revised polar model described in (1) of Scheme 1.24. Finally, Nelson also used this reactivity in the total synthesis of the apoptolidine C aglycone (59) [47]. In this case, silyl enol ether (60) participated in a highly stereoselective addition to chiral aldehyde (61) ((3) in Scheme 1.25), affording aldol (62) as a single diastereomer in 71% yield. A matched pairing of aldehyde and enolsilane facial biases acting in the transition state shown in Scheme 1.25 was invoked to explain such outstanding stereoselectivity [48].
Scheme 1.25 Use of 1,3-asymmetric induction imparted by chiral β-hydroxy aldehydes in Mukaiyama aldol reactions in the synthesis of natural products.
The aforementioned transformations show that the 1,3-anti asymmetric induction of protected β-hydroxy aldehydes can be very successful, but this control element should be used with caution because the diastereoselectivity drops when the steric bulk of the protecting group increases [49, 50]. Taking advantage of this trend, Yamamoto reported that the extremely bulky tris(trimethylsilyl)silyl group (TTMSS, (Me3Si)3Si), also known in the literature as the hypersilyl or super silyl group, confers to β-OTTMSS aldehydes an outstanding 1,3-syn asymmetric induction [51]. This silicon-protecting group is also remarkable because it permits acetaldehyde derived TTMSS enol ether to participate in highly diastereoselective Mukaiyama aldol additions to a large variety of aldehydes.2 Moreover, these additions can incorporate other TTMSS enol ethers in cascade aldol reactions with excellent levels of 1,2-Felkin and 1,3-syn-asymmetric induction. This is shown in the sequential addition of acetaldehyde- and acetophenone-derived TTMSS enol ethers to 2-phenylpropanal (Scheme 1.26) [51, 52]. On the basis of the open-chain model proposed by Evans ((1) in Scheme 1.24), the rationale for the observed 1,3-syn induction of β-OTTMSS aldehydes considers that the size of this silicon-protecting group now dictates that the carbonyl group is antiperiplanar to the OTTMSS substituent and far from the R1 group, which avoids unfavorable steric interactions. Therefore, conformation XVIII is responsible to the predominant formation of the 3,5-syn aldol diastereomer.
Scheme 1.26 Mukaiyama aldol reactions of tris(trimethylsilyl)silyl enolsilanes.
As discussed in Section 1.3.6, this chemistry combined with lithium-mediated aldol reactions from β-OTTMSS methyl ketones provides new entries to efficient synthesis of natural products.
1.2.4.3 Merged 1,2- and 1,3-Asymmetric Induction
The high asymmetric induction imparted by α- and β-substituted aldehydes documented in previous sections raises the prospect that there might exist intrinsic stereochemical relationships between both substituents that are either mutually reinforcing or opposing. This is particularly true for α-methyl β-alkoxy aldehydes.
Indeed, the stereochemical outcome of Mukaiyama aldol additions to these chiral aldehydes under nonchelation conditions can be easily interpreted by invoking the 1,2-Felkin and 1,3-anti asymmetric induction provided by the α-methyl and the β-oxygenated substituent, respectively. Hence, it is not surprising that the BF3-mediated additions of silyl enol ethers (42) to α-methyl-β-OPMB aldehydes (63) and (64) largely depend on the relative configuration of the aldehyde (Scheme 1.27). As the influences of the α-methyl and the β-OPMB substituents match in anti-substituted aldehyde (63), Felkin aldols (65) were virtually obtained as a single diastereomer, whereas syn-substituted aldehyde (64), in which those factors are opposing, gave mixtures of aldols (67) and (68) in variable diastereomeric ratios [53].
Scheme 1.27 Mukaiyama aldol additions to α-methyl β-alkoxy aldehydes.
Interestingly, additions to 64 showed a turnover in carbonyl face selectivity (from Felkin to anti-Felkin) on decreasing the size of the enolsilane group R. This implies that the β-stereocenter becomes the dominant control element in the reactions with sterically nondemanding enolsilanes. Moreover, a decrease in the solvent polarity produced an increase of anti-Felkin diastereomer (68), which means that 1,3-induction is enhanced relative to 1,2-induction in nonpolar media. After a seminal analysis of these stereochemical data, Evans concluded that 1,2-stereoinduction is found in those reactions proceeding through an antiperiplanar open transition state such as XX, while dominant 1,3-stereoinduction is manifest from a synclinal open transition state such as XXII. Then, a subtle balance of stereoelectronic effects and steric interactions between the Lewis acid and the R group of the nucleophile on transition states XX–XXIII represented in Scheme 1.28 determines the stereochemical outcome of these reactions [53].
Scheme 1.28Evans models for Mukaiyama aldol additions to syn α-methyl β-alkoxy aldehydes.
The consistently high stereocontrol provided by Mukaiyama aldol reactions of anti α-methyl β-oxygenated aldehydes has been successfully used in the synthesis of many natural products [54]. Evans reported that BF3-mediated vinylogous aldol addition of Chan's diene (52a) to chiral aldehyde (69) in toluene at − 90 °C gave β-hydroxy ketone (70), an advanced intermediate in the total synthesis of callipeltoside A (71), with excellent stereocontrol (dr > 95 : 5) in 88% yield ((1) in Scheme 1.29) [55]. In turn, Paterson developed a highly efficient coupling of chiral silyl enol ether (72) and elaborate α-methyl β-oxygenated aldehyde (73) to obtain aldol (74) as a single diastereomer in 91% yield, which was easily converted into preswinholide A (75) ((2) in Scheme 1.29) [56], the monomeric seco acid of swinholide A [57, 58]. Finally, the conversion of methyl ketone (76) into the corresponding silyl enol ether and subsequent BF3-mediated addition to aldehyde (77) in the presence of 4 Å molecular sieves allowed Roush to generate aldol (78), the penultimate precursor of bafilomycin A1 (79), with a high diastereoselectivity (dr > 95 : 5) in 72% yield ((3) in Scheme 1.29) [59].
Scheme 1.29 Use of asymmetric induction imparted by anti α-methyl β-oxygenated aldehydes in Mukaiyama aldol reactions in the synthesis of natural products.
Parallel Mukaiyama aldol additions to syn-α-methyl β-oxygenated aldehydes have also been applied to the synthesis of natural products, but the uncertainty of their stereochemical outcome and the poorer diastereoselectivities often provided by these aldehydes have restricted their use compared to their anti counterparts. Remarkably, most of the examples found in the literature about these transformations involve Felkin-like reactions leading to all syn aldols. For instance, Floreancig reported the preparation of Felkin aldol (82) in excellent yield as an inseparable 86 : 14 mixture of diastereomers by addition of sterically hindered silyl enol ether (80) to chiral syn-α-methyl β-silyloxy aldehyde (81) ((1) in Scheme 1.30) [60, 61]. Furthermore, Kalesse described one of the few examples of substrate-controlled Mukaiyama aldol reaction in which the 1,3-induction of the β-oxygenated stereocenter prevailed over the Felkin induction imparted by the α-stereocenter [62]. This engaged the silyl enol ether (42a) addition to syn-aldehyde (83), which afforded aldol (84) with excellent diastereoselectivity (dr 97 : 3) in 86% yield ((2) in Scheme 1.30) [63].
Scheme 1.30 Asymmetric induction imparted by syn α-methyl β-oxygenated aldehydes in Mukaiyama aldol reactions.
Following these analyses, π-facial selectivity of chiral aldehydes possessing α- and β-heteroatoms could be expected to arise from the summation of the inductions imparted by both substituents. Unfortunately, α,β-bisalkoxy aldehydes do not fulfill such expectations. Mukaiyama aldol additions to syn-α,β-bisalkoxy aldehydes under nonchelating conditions are too reliant on the hydroxyl protecting groups and the steric encumbrance of the enolsilane and usually proceed with poor stereocontrol. The same occurs to anti diastereomers, but anti-configured α-OBn β-OTBS aldehyde (85) exhibited uniformly high selectivities toward 3,4-anti aldols (86) irrespective of the steric hindrance of silyl enol ethers (42) ((1) in Scheme 1.31) [38]. A similar trend was observed for α-amino β-alkoxy aldehydes [64]. In this case, addition of silyl enol ether (42c) to anti α-N-BnTs β-OTBS aldehyde (87) afforded 3,4-anti aldol (88) as a single diastereomer in 92% yield ((2) in Scheme 1.31).
Scheme 1.31 Asymmetric induction imparted by chiral aldehydes possessing α- and β-heteroatoms in Mukaiyama aldol reactions.
The highly diastereoselective aldol reaction of silyl enol ether (ent-38) and anti-aldehyde (89) affording anti aldol (90) is consistent with these features ((1) in Scheme 1.32). Moreover, alternative patterns occasionally proceed with excellent levels of diastereoselectivity. For instance, Paterson reported that silyl enol ether (42a) and syn-aldehyde (91) participated in a highly diastereoselective reaction (dr 93 : 7) toward all syn-aldol (92), thus indicating that the steric effect of the large alkyl group overrode any electronic stereocontrol from the oxygenated substituents ((2) in Scheme 1.32) [65, 66]. In spite of these successful examples, most of the stereocontrolled aldol reactions of α, β-bisalkoxy aldehydes rely on the use of lithium enolates (Section 1.3.5).
Scheme 1.32 Mukaiyama aldol reactions involving α,β-bisalkoxy aldehydes.
The stereochemical outcome of all these Mukaiyama aldol reactions involving α, β-disubstituted aldehydes can be dramatically affected if they are carried out under chelating conditions. Particularly, Evans established that the choice of the Lewis acid was crucial to the control of the additions to syn α-methyl β-OPG (PG: Bn, TBS) aldehydes (93) (Scheme 1.33) [67].3 Thus, BF3-mediated additions of silyl enol ether (42c) to these aldehydes provided Felkin aldols (94) in excellent diastereomeric ratios (dr > 96 : 4), whereas opposite anti-Felkin aldols (95) were obtained with the same level of diastereoselectivity by using Me2AlCl.4 Further theoretical and spectroscopic studies revealed that these transformations proceed through a cationic dimethylaluminum chelate that preferentially adopts a boat conformation and directs the attack of the nucleophile to the Si face of the C—O bond (Scheme 1.33). From a general point of view, these results demonstrate the exceptional chelating ability of this aluminum Lewis acid and expand the stereochemical control provided by these aldehydes [68, 69].
Scheme 1.33 Influence of Lewis acids on the stereochemical outcome of Mukaiyama aldol reactions of α-methyl β-alkoxy aldehydes.
1.2.5 Chiral Lewis Acids
Not surprisingly, the crucial role played by Lewis acids in Mukaiyama aldol reactions has stimulated the search for chiral ligands to dictate the stereochemical outcome of these transformations. The resultant chiral Lewis acids must increase the electrophilicity of aldehydes, create a suitable asymmetric environment around the carbonyl bond, and, as far as possible, facilitate catalytic turnover at the same time [5]. The following section describes how the intense efforts directed to this challenging objective have already provided highly enantioselective approaches, paying particular attention to those chiral Lewis acids used in the synthesis of natural products.
Mukaiyama and Kobayashi reported the first chiral complexes to provide stereocontrolled acetate Mukaiyama aldol reactions. These initially consisted of three pieces: Sn(OTf)2, an optically active diamine, and a tin(IV) additive. Thus, stoichiometric amounts of Sn(OTf)2, proline-derived diamines (96), and tributyltin fluoride promoted highly enantioselective additions of silyl ketene S,O-acetal (36a) to representative aldehydes, leading to the corresponding β-hydroxy thioesters (97) (Scheme 1.34) [70]. Application of these conditions to the TBS ketene acetal from benzyl acetate furnished similar results [71].
Scheme 1.34 Asymmetric Mukaiyama aldol reactions promoted by stoichiometric amounts of Sn(OTf)2 and proline-derived diamines.
Remarkably, these additions were also successful when using catalytic amounts of Sn(OTf)2 and diamine (96b) without tributyltin fluoride provided that the reaction was carried out in propionitrile, and the enolsilane and the aldehyde were both added slowly to the reaction mixture [72]. The power of this methodology was demonstrated on the addition of 36a to chiral α-OTBS aldehydes (98) and (ent-98), in which the C3 stereocenter of aldols (99) and (100) was controlled by the diamine (96) regardless of the inherent diastereofacial preference of the chiral aldehydes (Scheme 1.35) [73, 74].
Scheme 1.35 Asymmetric Mukaiyama aldol reactions promoted by catalytic amounts of Sn(OTf)2 and proline-derived diamines.
Mechanistic analyses of this reaction suggested that such excellent stereocontrol might arise from the approach of the enolsilane to the less hindered face of the aldehyde in the highly ordered complex shown in Scheme 1.36 [75]. Regarding the catalytic turnover, the metal exchange reaction with TMSOTf to regenerate the catalyst was identified as a crucial step, because TMSOTf can promote an alternative achiral route (Scheme 1.36) that erodes the enantioselectivity if the silyl transfer is slow (kturnover ≈ kachiral). Thus, the need to minimize this side reaction led to the use of a more polar solvent such as propionitrile, which increases kturnover, and to the slow addition of both the enolsilane and the aldehyde to the reaction mixture.
Scheme 1.36 Mechanism for the Mulaiyama aldol reaction catalyzed by Sn(OTf)2 and a proline-derived diamine.
This methodology showed that catalytic asymmetric Mukaiyama aldol reactions were possible and could be used successfully in organic synthesis, but it suffered from the difficulty of handling Sn(OTf)2 [76] and the need for the reagents to be added to the reaction mixture slowly.
In the early 1990s, some chiral catalysts derived from the easily available BINOL ligand emerged as candidates to overcome such limitations. Mukaiyama was the first to use a BINOL-derived titanium–oxo complex to catalyze the addition of silyl ketene S,O-acetals to a limited number of aldehydes with modest stereocontrol [77]. This was followed by Mikami's findings on highly enantioselective ene and aldol reactions catalyzed by BINOL–titanium complex (101). Remarkably, addition of silyl enol ether (26a) to methyl glyoxylate afforded α-hydroxy ester (102) as a single enantiomer ((1) in Scheme 1.37) [78], whereas aldol reactions of silyl ketene S,O-acetals (36) and a wide array of aldehydes furnished thioesters (103) in good to high yields and enantiomeric ratios up to 98 : 2 ((2) in Scheme 1.37) [79].
Scheme 1.37 Asymmetric Mukaiyama aldol reactions catalyzed by BINOL-TiCl2.
Crossover experiments and analyses of the stereochemical outcome of related reactions led to the suggestion of a cyclic six-membered transition state for these transformations, wherein the oxygen of the activated carbonyl bond interacts with the transferring group, H or SiR3 for ene or aldol reactions, respectively (Scheme 1.38). Furthermore, the enantioselectivity was further improved by the addition of achiral ligands or by more bulky silicon groups, which suggests a more complex mechanism [80]. Irrespective of the mechanism, the synthetic potential of both transformations was demonstrated in two-directional ene processes [81] and aldol reactions involving chiral aldehydes by using a catalyst derived from BINOL and TiCl2(i-PrO)2 [82].
Scheme 1.38 Cyclic transition states for ene and aldol reactions.
Keck reported a similar methodology in which the titanium catalyst was prepared by mixing BINOL ligand and Ti(i-PrO)4 [83]. The stereocontrol was again excellent, but Keck warned about the unknown structure of the catalytic species and emphasized the pronounced sensitivity of these reactions to BINOL–metal stoichiometry and the presence of other additives. Indeed, the catalytic species can be prepared according to several protocols adapted to different nucleophiles. Therefore, these have been applied to reactions from silyl ketene S,O-acetal (36b) [83], diene (104) [84, 85], and vinyl ketene S,O-acetal (105) [86] leading to β-hydroxy thioesters (106), dihydropyrones (107), and β-hydroxy α, β-unsaturated thioesters (108) in a highly enantioselective manner ((1–3) in Scheme 1.39).
Scheme 1.39 Asymmetric Mukaiyama aldol reactions catalyzed by BINOL/Ti(i-PrO)4.
A thorough analysis and optimization of these transformations revealed the nonlinear effects and the dramatic influence of some additives [87, 88], which confirmed Keck's comments about the sensitivity of these reactions to structural and experimental modifications [89]. In spite of these limitations, these catalysts are an important tool for the stereocontrolled synthesis of natural products [90]. This is illustrated by the catalytic addition of Chan's diene (52b) to cinnamaldehyde affording aldol (109), an intermediate in the total synthesis of (+)-cryptofolione (110) described by Meshram (Scheme 1.40) [91].
Scheme 1.40 Synthesis of (+)-cryptofolione.
At the same time that Mikami and Keck developed their BINOL-derived titanium catalysts, Carreira reported that mononuclear Ti(IV) complex (111a), generated from a chiral tridentate Schiff base, Ti(i-PrO)4, and 3,5-di-tert-butylsalicylic acid, triggered extremely efficient reactions of silyl ketene acetals (112) and a wide range of aldehydes [92] including α, β-ynals [93] to yield β-hydroxy esters (113) with enantiomeric ratios up to 99 : 1 ((1) in Scheme 1.41). The stereocontrolled synthesis of ester (115) by addition of silyl ketene acetal (112a) to chiral aldehyde (114) in the total synthesis of roflamycoin (116), reported by Rychnovsky ((2) in Scheme 1.41) [94], is a convincing demonstration of the potential of this methodology [95]. This method was also applied to O-silyl dienolate (117), which yielded carbonyl adducts (118) in a stereocontrolled manner ((3) in Scheme 1.41) [96]. Furthermore, a parent chiral Ti(IV) complex (111b) prepared by mixing the chiral tridentate Schiff base and Ti(i-PrO)4 (the structure of the active catalyst has not been determined yet) promoted the enantioselective addition of 2-methylpropene to aldehydes leading to β-hydroxy methyl ketones (119) ((4) in Scheme 1.41) [97].
Scheme 1.41 Asymmetric Mukaiyama aldol reactions catalyzed by Carreira's titanium complex.
The success of all these transformations is largely due to the effective transfer of the silyl group. As represented in Scheme 1.42, this transfer is understood to occur intramolecularly through an intermediate XXIV, wherein the ligand associated with the metal complex serves as a shuttle, transiently undergoing silylation. In this context, the replacement of isopropoxide groups by salicylic acid in catalyst (111) became crucial, as it facilitated the silyl transfer, which finally produced a strong increase of yields, enantioselectivities, and catalytic turnover.
Scheme 1.42 Mechanism for the asymmetric Mukaiyama aldol reaction catalyzed by Carreira's titanium complex.
The prominence of boron Lewis acids in metal-enolate-mediated aldol reactions (Section 1.3) often overshadows their contribution to the development of asymmetric Mukaiyama aldol counterparts. Nevertheless, oxazaborolidinone complexes readily prepared from α-amino acids are among the most successful chiral Lewis acids for highly stereocontrolled transformations [98].
Introduced by Kiyooka, valine-derived N-tosyl oxazaborolidinone (120) was successfully engaged in stoichiometric Mukaiyama aldol reactions [99, 100]. Its ability as a chiral promoter was demonstrated in synthetic studies involving a wide array of enolsilanes derived from phenyl acetate, methyl acetoacetate, and methyl ketones (Scheme 1.43) [101, 102]. Remarkably, 120 facilitated the addition of silyl enol ether (121) to chiral aldehyde (122) and the subsequent syn reduction of the resulting aldol adduct to give diol (123) with excellent diastereoselectivity in 64% yield.
Scheme 1.43 Asymmetric Mukaiyama aldol reactions catalyzed by a valine-derived N-tosyl oxazaborolidinone.
In turn, Corey reported that Mukaiyama aldol additions of silyl enol ethers (42) to representative aldehydes in the presence of catalytic amounts of tryptophan-derived chiral borane complex (124) furnished β-hydroxy ketones (125) enantioselectively ((1) in Scheme 1.44) [103].5 In a fine example of tandem catalysis, Carreira used this complex to construct the azabicyclononane core common to a family of biomolecules [104]. Thus, it catalyzed the Diels–Alder reaction of furan and bromoacrolein followed, on consumption of the educts, by aldol addition of silyl ketene acetal (112b), which led to the stereoselective formation of β-hydroxy ester (126) ((2) in Scheme 1.44) [105].
Scheme 1.44 Asymmetric Mukaiyama aldol reactions catalyzed by a tryptophan-derived chiral borane complex.
Mechanistic studies on these and other reactions and X-ray crystallographic evidence led to the suggestion that the stereocontrol achieved by these transformations arose from the assemblies shown in Scheme 1.45, wherein a hydrogen bond between formyl and oxygen came up as a crucial stereocontrol element [106]. Thus, pyramidalization of the sulfonamide due to a gearing interaction with the isopropyl substituent in Kiyooka's complex XXV orients the sulfonamide residue to shield the Re face of the aldehyde, which impels the nucleophile to approach the opposite face. Conversely, attractive π − π interaction in Corey's complex XXVI blocks the Si face of the aldehyde and favors the attack of the nucleophile on the less hindered Re face.
Scheme 1.45 Transition states for asymmetric Mukaiyama aldol reactions catalyzed by oxazaborolidinones.
On the basis of these and other precedents [107], Kalesse reported that tryptophan-derived B-aryl oxazaborolidinones (127) and (128) underwent enantioselective vinylogous Mukaiyama aldol reactions from dienes (129) and (130), furnishing δ-hydroxy esters (131) or aldehydes (132) in high yields under stoichiometric conditions ((1) and (2) in Scheme 1.46). Importantly, the latter is the first asymmetric vinylogous Mukaiyama aldol reaction with aldehyde-derived silyl dienol ethers. Moreover, B-phenyl oxazaborolidinone (127) was successfully used by Fürstner to assemble silyl enol ether (133) and aldehyde (134) in the penultimate step of the total synthesis of lactimidomycin (135) ((3) in Scheme 1.46) [108]. In turn, Harada also found that an allo-threonine-derived B-phenyl oxazaborolidinone catalyzed the enantioselective addition of dimethylsilyl ketene S,O-acetals to acetophenone [109].
Scheme 1.46 Asymmetric Mukaiyama aldol reactions catalyzed by tryptophan-derived B-aryl oxazaborolidinones.
In addition to these Lewis acids, Evans disclosed that Cu(II), Sn(II), and Sc(III) complexes of chiral pyridine-bisoxazoline (pybox) and bisoxazoline (box) ligands catalyze the asymmetric Mukaiyama aldol reactions of chelating electrophiles. For instance, low catalyst loadings of Cu(II) pybox–complex (136) were enough to trigger the addition of several nucleophiles to benzyloxyacetaldehyde, affording the corresponding aldol adducts (137–140) in high yields and excellent stereocontrol (Scheme 1.47) [110].
Scheme 1.47 Asymmetric Mukaiyama aldol reactions catalyzed by a pybox copper complex.
The proposed catalytic cycle for these aldol reactions is outlined in Scheme 1.48. Coordination of the aldehyde to the Cu(II) center produces complex XXVII, which undergoes the nucleophilic addition to produce a copper aldolate XXVIII. Importantly, both intra- and intermolecular fast silyl transfers form XXIX and subsequent decomplexation affords the aldol adduct and regenerates catalyst (136). In turn, a stereochemical model XXX containing a chelated α-OBn aldehyde at the metal center, which forms a square pyramidal copper intermediate, accounts for the direction of the induction observed.
Scheme 1.48 Mechanism for asymmetric Mukaiyama aldol reactions catalyzed by a pybox copper complex.
Unfortunately, only benzyloxy-like acetaldehydes participated in this highly enantioselective reaction. Other aldehydes capable of engaging in five-membered chelated complexes such as glyoxalate esters required Sn(II)-box or Sc(III)-pybox complexes (141a and 142a, respectively) to react with silyl ketene S,O-acetals with similar levels of enantioselectivity ((1) and (2) in Scheme 1.49) [111, 112]. Therefore, other enolsilanes were surveyed in an effort to expand the scope of this glyoxalate aldol process. It was found that the addition of silyl enol ethers (143) catalyzed by Sc(III)-pybox (142b) furnished α-hydroxy-γ-keto esters (144) in high yields and excellent enantioselectivity, provided that TMSCl was added to facilitate catalyst turnover ((3) in Scheme 1.49). Importantly, t-Bu-catalyst (142b) afforded the opposite direction of induction to that of 142a. Steric factors within the chiral pocket affecting the location of the aldehyde carbonyl in the complex have been proposed to explain this different behavior. Furthermore, Evans reported in a brilliant study that Sn(II)-box (141b) catalyst promoted highly enantioselective additions of silyl ketene S,O-acetals to pyruvate esters affording functionalized hydroxysuccinate derivatives (145) with an absolute stereocontrol ((4) in Scheme 1.49) [113]. The scope of this reaction was further extended to include nucleophiles such as silyl enol ethers from methyl ketones and other electrophiles capable of providing five-membered chelated complexes such as 1,2-diketones. In spite of such advances, these transformations lack generality and can only be applied to a selected set of electrophiles, although their outstanding levels of enantioselectivity have allowed them to be widely used in the synthesis of natural products [114].
Scheme 1.49 Asymmetric Mukaiyama aldol reactions catalyzed by box and pybox complexes.
Beyond the results themselves, these studies had a profound impact on asymmetric reactions, because they introduced new concepts and chiral ligands suitable for Lewis-acid-mediated transformations based on aldehydes or ketones with chelating functional groups. Therefore, it is not surprising that a large amount of work has been devoted to surveying alternative box- and pybox-like ligands or to identifying other metals to catalyze such reactions [115–117]. Furthermore, they also inspired the development of new catalysts, such as C1-symmetric aminosulfoximines reported by Bolm, which undergo Cu(II)-catalyzed reactions with pyruvate esters [118], or a peptide bearing a Schiff base disclosed by Hoveyda, which promotes highly successful and experimental friendly AgF2-mediated stereocontrolled additions to a wide array of α-keto esters [119].
Regarding the use of copper-derived Lewis acids, Carreira devised a different way of taking advantage of the resultant chiral catalysts [120]. This involved the addition of O-silyl dienolate (117) to α, β-unsaturated and aromatic aldehydes catalyzed by a chiral copper complex generated in situ by mixing (S)-Tol-BINAP (146), Cu(OTf)2, and (Bu4N)Ph3SiF2 (TBAT), which afforded aldol adducts (147) with enantiomeric ratios up to 98 : 2 ((1) in Scheme 1.50). In turn, Campagne expanded the scope of this methodology to methyl ketones and diene (148), obtaining lactones (149) in modest to high yields in stereocontrolled way ((2) in Scheme 1.50) [121]. This reliable methodology has been often used in the synthesis of natural products. For instance, Scheidt used the diastereoselective reaction of 117 and chiral aldehyde (150) to prepare aldol (151), an important fragment of the total synthesis of (−)-okilactomycin (152) ((3) in Scheme 1.50) [122, 123].
Scheme 1.50 Asymmetric Mukaiyama aldol reactions catalyzed by a Tol-BINAP copper complex.
Importantly, other Cu(I) sources served equally well as Cu(OTf)2. This result, along with the known reduction of Cu(II) to Cu(I) by enolsilanes, suggested that a Cu(I) complex was the catalytic species. Further spectroscopic studies confirmed this hypothesis and also supported a mechanism in which a copper enolate acted as the reactive species ((1) in Scheme 1.51). Nevertheless, studies on related additions from dienes such as 148 led Campagne to point out that this model did not account for the origin of the enantioselectivity because the chiral copper center is far away from the aldehyde. Instead, the reaction might proceed through an initial nonselective α-aldol, followed by a retroaldol step, producing a C-enolate that would evolve toward an allyl-copper intermediate, which would finally undergo an asymmetric allylation ((2) in Scheme 1.51) [124].
Scheme 1.51 Proposed mechanisms for Mukaiyama aldol reactions catalyzed by Tol-BINAP copper complexes.
Finally, some studies have recently focused on the use of chiral Brønsted acids and other protic catalysts to control the stereochemical outcome of these reactions. Rawal described the first examples of hydrogen-bond-mediated enantioselective Mukaiyama acetate aldol reactions, which involved additions of O-silyl dienolates to reactive aldehydes in the presence of diols as catalysts [125, 126]. Furthermore, Yamamoto found that catalytic amounts of N-triflylthiophosphoramide (153) promoted stereocontrolled additions of silyl enol ethers (154) from aryl methyl ketones to α, β-unsaturated and aromatic aldehydes, affording the corresponding β-hydroxy ketones (155) in high yields (Scheme 1.52) [127]. Importantly, mechanistic studies revealed that the actual catalyst was 153 itself rather than the silylated Brønsted acid.
Scheme 1.52 Asymmetric Mukaiyama aldol reactions catalyzed by a chiral N-triflylthiophosphoramide.
In turn, List reported that chiral disulfonimide (156) was an efficient catalyst for different Mukaiyama aldol reactions. Indeed, tiny amounts of this chiral Brønsted acid triggered additions of silyl ketene acetal (27c) to nonenolizable aldehydes, producing β-OTBS isopropyl esters (157) in high yields and enantioselectivities ((1) in Scheme 1.53) [128]. This chemistry was further extended to vinylogous and bisvinylogous additions of (158a) and (159) to α, β-unsaturated and aromatic aldehydes, which afforded the corresponding methyl esters (160) and (161) in a regio and stereocontrolled manner ((2) and (3) in Scheme 1.53) [129]. The mechanism of these transformations is unknown, but it was proposed that 156 might be first silylated by the ketene acetal, providing an N-silyl disulfonimide that could activate the aldehyde through O-silylation. Thus, the asymmetric induction would occur by stereochemical communication within an ion pair consisting of the disulfonimide anion and the silylated oxonium cation. Irrespective of such mechanistic considerations, these findings have paved the way for the development of a new sort of transformations based on the appropriate choice of chiral hydrogen-bond donors [130].
Scheme 1.53 Asymmetric Mukaiyama aldol reactions catalyzed by a chiral disulfonimide.
1.2.6 Chiral Lewis Bases
In addition to the above-mentioned methodologies based on the increase of the electrophilicity of aldehydes by binding Lewis or Brønsted acids to the carbonyl group, Denmark developed a variant of the Mukaiyama aldol reaction that takes advantage of the Lewis acidic silicon atom of trichlorosilyl enolates. Thereby, association of these enolates with Lewis bases gives rise to a hypervalent silicon moiety characterized by enhanced Lewis acidity. Thus, such a reacting species is poised for activation of the carbonyl component, which allows it to participate in highly efficient aldol reactions (Scheme 1.2) [9].
Systematic investigation on aldol reactions of trichlorosilyl enolates revealed that chiral phosphoramide (162) triggered highly enantioselective additions of methyl ketone-derived enolates (163) to branched aliphatic, α, β-unsaturated, and aromatic aldehydes, affording β-hydroxy ketones (164) in high yields ((1) in Scheme 1.54). Importantly, the Hg(II)-catalyzed transilylation of TMS enol ethers (42) with SiCl4 provided an entry to a one-pot method for the generation and reaction of these enolates with similar yields and enantioselectivities ((2) in Scheme 1.54) [131]. Echavarren used this method to prepare β-hydroxy ketone (165), an intermediate in the enantioselective synthesis of (−)-englerins A and B (166) ((3) in Scheme 1.54) [132].
Scheme 1.54 Asymmetric Mukaiyama aldol reactions catalyzed by chiral phosphoramides.
This method was further tested on chiral substrates with different results. For instance, the additions of trichlorosilyl enolate from lactate-derived silyl enol ether (167) to benzaldehyde catalyzed by 162 or ent-162 produced the same 1,4-syn aldol (168) in both cases ((1) in Scheme 1.55). Conversely, parallel aldol additions of trichlorosilyl enolates from β-silyloxy silyl enol ethers (169) and (170) produced the corresponding aldol adducts (171–174) depending on the phosphoramide ((2) and (3) in Scheme 1.55) [133].
Scheme 1.55 Mukaiyama aldol reactions of chiral trichlorosilyl enolates catalyzed by chiral phosphoramides.
The mechanism of all these reactions is rather complex. Indeed, exhaustive studies established that the process takes place via the simultaneous operation of two mechanistic pathways involving one or two phosphoramides bound to the silicon atom in cationic complexes XXXI and XXXII (Scheme 1.56). Their subsequent association with the aldehyde leads to a reversible albeit unfavorable formation of an activated complex, which evolves through cyclic transition states XXXIII or XXXIV. The rate of this cycle determines the turnover as well as the stereochemical outcome of the aldol reaction [134]. Thus, the achievement of highly stereocontrolled transformations requires that the aldol reaction mainly proceeds through one of these competitive pathways.
Scheme 1.56 Mechanistic profile adapted from the original proposed for the trichlorosilyl enolate from cyclohexanone.
A second type of Lewis base catalysis involves activation of weakly acidic SiCl4 by binding of strong Lewis basic chiral phosphoramide (175), which leads to the formation of a chiral Lewis acid in situ. This species has been shown to be a competent catalyst for Mukaiyama aldol additions of achiral and chiral silyl enol ethers to α, β-unsaturated and aromatic aldehydes (Scheme 1.57) [133, 135].
Scheme 1.57 SiCl4-mediated Mukaiyama aldol reactions of silyl enol ethers catalyzed by chiral phosphoramides.
Furthermore, these procedures were also successful for the additions of silyl ketene acetal (27a) and vinylogous systems (158b) and (176) to a broad scope of aldehydes, affording β- and δ-hydroxy esters (177–179) in high yields and exceptional levels of regio- and enantioselectivity, which expanded the synthetic potential of this methodology (Scheme 1.58) [136–138].
Scheme 1.58 SiCl4-mediated Mukaiyama aldol reactions catalyzed by chiral phosphoramides.
Although the origin of the observed enantioselectivity is still unclear, spectroscopic and mechanistic studies of these reactions suggested that the catalytic cycle involved a highly electrophilic, phosphoramide-bound silyl cation XXXV (Scheme 1.59). This species could then bind the aldehyde to form complex XXXVI, which would undergo the addition of the enolsilane through an open transition state. After cleavage of the silyl group and dissociation of the catalyst, the resultant aldolate XXXVII would give a trichlorosilyl ether (Scheme 1.59) [136, 139].
Scheme 1.59 Mechanism for the SiCl4-asymmetric Mukaiyama aldol reactions catalyzed by chiral phosphoramides.
Finally, Denmark also reported that bis N-oxide (180), a Lewis base, catalyzed the aldol addition of methyl trichlorosilyl ketene acetal to a wide range of nonactivated ketones (Scheme 1.60). The resultant β-tert-hydroxy esters were obtained in excellent yields with enantioselectivities highly dependent on the structure of the ketone [140]. From a mechanistic point of view, this reaction is understood to proceed through a six-membered boatlike transition state organized around a cationic silicon center, in a way similar to the aldol additions to aldehydes described in Scheme 1.56.
Scheme 1.60 Asymmetric Mukaiyama aldol reactions catalyzed by a chiral bis N-oxide.
1.3 Metal Enolates
1.3.1 Concept and Mechanism
The enolization of a parent carbonyl compound with a strong base, a suitable combination of a Lewis acid and a tertiary amine, or by simple transmetallation affords metal enolates from nearly every element of the periodic table [141], which suggests that any element could participate in asymmetric aldol reactions. Nevertheless, only a narrow group of elements including lithium, boron, titanium, or tin fulfill the required conditions to undergo stereocontrolled aldol transformations. In essence, the configurationally stable O-enolates of these elements allow the aldol reactions to proceed through cyclic and highly organized transition states. Such transition states provide the appropriate environment to control the new stereocenters provided that the appropriate chiral elements are placed on the substrate, the aldehyde, or the ligands bound to the metal. Unfortunately, acetate aldol reactions mediated by such intermediates can evolve through different six-membered cyclic transition states XXXVIII–XLI (Scheme 1.61), which hampers the proper differentiation of the two faces of the carbonyl bond by the unsubstituted enolate.
Scheme 1.61 Metal-enolate-mediated acetate aldol reactions.
