Directed Selectivity in Organic Synthesis - Tanja Gaich - E-Book

Directed Selectivity in Organic Synthesis E-Book

Tanja Gaich

0,0
55,99 €

-100%
Sammeln Sie Punkte in unserem Gutscheinprogramm und kaufen Sie E-Books und Hörbücher mit bis zu 100% Rabatt.
Mehr erfahren.
Beschreibung

Bringing together examples that until now were often hidden and widely spread throughout the original literature, this textbook shows how to use the correct reagents, conditions or reaction sequences to have access to all possible stereoisomers when beginning synthesis with only a single starting material.

Adopting a didactic approach, the authors have chosen general and simple examples throughout the book so that these reactions can be transferred easily to other reaction types.

While of major interest to master and PhD students alike, this book is also a source of valuable information for organic chemists in both academia and industry.

Additional material for lectures at www.wiley-vch.de/textbooks

Sie lesen das E-Book in den Legimi-Apps auf:

Android
iOS
von Legimi
zertifizierten E-Readern

Seitenzahl: 387

Veröffentlichungsjahr: 2014

Bewertungen
0,0
0
0
0
0
0
Mehr Informationen
Mehr Informationen
Legimi prüft nicht, ob Rezensionen von Nutzern stammen, die den betreffenden Titel tatsächlich gekauft oder gelesen/gehört haben. Wir entfernen aber gefälschte Rezensionen.



Table of Contents

Related Titles

Title Page

Copyright

Preface

Acknowledgement

Chapter 1: General Methods to Direct Selectivity

1.1 Chemoselectivity

1.2 Regioselectivity

1.3 Stereoselectivity

1.4 Enatioselectivity

References

Chapter 2: Directed Selectivity: Acetylenes and Alkenes

2.1 Acetylenes

2.2 Alkenes

References

Chapter 3: Directed Selectivity with Carbonyl Derivatives

3.1 Carbonyl Compounds as Electrophiles

3.2 Conjugate Addition to Carbonyl Compounds

3.3 Enolate Derivatives

3.4 Miscellaneous

References

Chapter 4: Selectivity at sp3 – Centers and Heteroatoms

4.1 Epoxides

4.2 Acetals

4.3 Ethers

4.4 Alcohols and Phenols

4.5 Sulfur Compounds

4.6 Amines

4.7 Halogen Compounds

References

Index

Related Titles

Christmann, M., Bräse, S. (eds.)

Asymmetric Synthesis II

More Methods and Applications

2012

Print ISBN: 978-3-527-32900-7, also available in electronic formats

Nicolaou, K.C., Chen, J.S.

Classics in Total Synthesis III

Further Targets, Strategies, Methods

2011

Print ISBN: 978-3-527-32957-1

Sierra, M.A., de la Torre, M., Cossio, F.P.

More Dead Ends and Detours

En Route to Successful Total Synthesis

2013

Print ISBN: 978-3-527-32976-2, also available in electronic formats

Terrier, F.

Modern Nucleophilic Aromatic Substitution

2013

Print ISBN: 978-3-527-31861-2, also available in electronic formats

The Authors

Dr. Tanja Gaich

Leibnitz Univ. Hannover

Inst. of Organic Chemistry

Schneiderberg 1

30167 Hannover

Germany

Prof. Ekkehard Winterfeldt

Leibnitz Univ. Hannover

Inst. of Organic Chemistry

Schneiderberg 1

30167 Hannover

Germany

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>.

© 2014 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: 978-3-527-33375-2

ePDF ISBN: 978-3-527-66730-7

ePub ISBN: 978-3-527-66729-1

Mobi ISBN: 978-3-527-66728-4

Preface

Whenever the future goals of chemistry, and particularly synthetic chemistry or chemical production, are being discussed, the term sustainability will certainly play a major role in all arguments.

Although quite a bundle of aspects contribute to this important but still quite distant aim, selectivity is, doubtless, the most obvious and most important challenge in the field of synthetic chemistry.

Only with very high, reliable, and easy-to-apply and easy-to-manipulate selectivity will we ever succeed in gaining a high degree of efficiency and flexibility, which is “Conditio Sine Qua Non” to achieve sustainability.

In accordance with this, research in synthetic chemistry has concentrated very successfully in the last few decades on all aspects of selectivity – on highly selective reagents and catalysts as well as on various techniques – to manipulate the structure and the conformation of substrates and to prepare them for a reliably directed approach with all kinds of reaction partners.

This resulted in an impressive arsenal of highly selective reagents and catalysts and also disclosed a manifold of methods to achieve a high degree of chemoselectivity, regioselectivity, diastereoselectivity, and enantioselectivity.

All this has been compiled in review articles and books and can easily be obtained from data banks and synthesis programs.

We believe, therefore, that a general treatment of selectivity is by far beyond the frame of a publication such as this and that we would just duplicate all the information that can easily be collected from electronic data anyway. If it comes to directed selectivity, however, retrieval is not that easy.

We use the term directed selectivity in those cases where a set of transformations will lead to both or all possible sterically defined entities from one single starting material by simply changing reagents, reaction conditions, or the reaction sequence.

This approach is of particular value if the chosen starting material such as a chiral pool compound or an industrial bulk product is cheap or easily available, and if one is aiming at libraries of isomers or stereoisomers to study biological activities or other properties in a quite special group of compounds.

While aiming at directed selectivity, the search for relevant data can sometimes be quite frustrating, since the results of interest may be “hidden” in a by-product, by-pass, or dead end of a synthetic investigation, in a passage of product optimization not mentioned in the abstract, in a mechanistic exercise, or in the optimization of a catalyst.

From various sources of this kind, we have collected numerous references over a period of 15–20 years as contributions to workshops for synthetic chemists.

The presentation starts with a compilation of the most important methods to manipulate the approach and the interaction of reagents with substrates.

This can be done either by changing the constitution or the conformation of the substrate, by varying the electronic nature or space demand of the reagent or catalyst, by changing the reaction mechanism, or by kinetic versus thermodynamic control.

In the second part, we present applications of these options in selective transformations of various synthetically useful functional groups that are generally present in a majority of starting materials or intermediates.

Whenever possible, quite simple and hopefully easy-to-generalize examples are selected; we have rarely included highly developed and very specialized structures, such as advanced intermediates from a total synthesis, since in these quite unique cases transfer to standard type molecules could create problems.

There will also be no general discussion of the reactions presented and no treatment of mechanistic proposals.

Since there is general agreement that in a synthetic venture one should make sure to introduce the correct configuration as soon as possible, reliable and highly flexible selective transformations are mandatory at this early stage of a synthetic enterprise.

Finally, it has to be stressed that we have not made any effort to reach complete coverage. In our selection of topics and examples, we were strongly guided by our own experiences in selective synthesis and our personal judgment on the most annoying difficulties in this field and on the possibilities to solve these problems.

We hope that this very personal selection of examples will properly describe the key challenge in directed selectivity, and we very much apologize to all colleagues who were not cited although they certainly have contributed substantially to this field too.

Leipzig

Tanja Gaich

April 2013

Ekkehard Winterfeldt

Acknowledgement

We wholeheartedly thank the Institute of Organic Chemistry of the Leibniz University of Hannover for their endorsement. Especially the very strong and important technical contribution of Christine Bartetzko and Sabine Ohlrogge to the preparation of the manuscript is most gratefully acknowledged.

Chapter 1

General Methods to Direct Selectivity

In the first chapter, we shall focus on the different modes of selectivity dealt with in organic synthesis and we shall describe the most important general methods to direct selectivity in these fields.

1.1 Chemoselectivity

The most obvious area that has already been intensely treated over many years is chemoselectivity [1].

The majority of the problems here have been solved to date, mainly with the help of protecting groups.

This is a broad field, but since it has been expertly and comprehensively covered in books [2] and review articles [3], we shall not engage in the same here.

In addition, there is a tendency in the last years to leave protecting groups altogether [4], since their removal may sometimes create problems at a later stage and since they mean additional steps, it translates into additional time and efforts.

Consequently, we nowadays aim at chemoselectivity without protecting groups.

A very simple solution is to hide the functional group in a reversible manner as, for instance, with the enolate of a carbonyl group [5].

While the higher δ – character of the keto group in ketoester 2 allows for mild borohydride reduction to yield hydroxyester 1, this may lead to preferential enolate formation followed by selective hydride reduction of the ester group to generate hydroxyketone 3.

As polarization and enolization of carbonyl groups are the crucial steps in these efforts, one is not surprised that oxophilic countercations such as aluminum and magnesium are particularly helpful and that they manage to trigger the in situ enolate formation.

This is nicely demonstrated with the selective diisobutylaluminum hydride (DIBAL)-reduction of β-dicarbonyl compound 4 [6].

Probably the oxophilic aluminum compound attacks the carbonyl groups to form 5, which is then reduced to enolate 7. As long as this enolate is not quenched by protonation, one could continue with other transformations in a molecule of this type without touching the 1,3-dicarbonyl moiety. As predicted, this type of enolate formation can also be exercised with magnesium as the countercation, and as an example one notices the dimerization of cyanoacetate to form the β-dicarbonyl system 10 [7].

While deprotonation with sodium methoxide leads to nitrile attack forming enamine 9, the employment of magnesium methoxide favors chelation of the Claisen intermediate, giving rise to the 1,3-dicarbonyl compound 10.

In situ manipulation also plays a vital role in the selective reduction of ketoaldehyde 11 in the presence of cerium trichloride [8]

as well as in the allene formation from the butynediol derivatives 13 [9].

While in all these cases we dealt with complexation of the substrate to modify the electronic behavior, one may also use complexation to enhance or to reduce the reactivity of reagents [10].

Typical and very well-established borohydride complexes range from cyanoborohydride 15 via the various alkoxy compounds 16 to tris-acetoxyborohydride 17 and tris-alkylborohydride 19.

Very similar to the trisacetoxy compound 17, which is simply obtained by dissolving sodium borohydride in acetic acid, the tris-tert-butoxy-alanate complex 18 is formed also on treatment of lithium alanate with tert-butyl alcohol.

In both cases, only three hydride anions are displaced, leading in the case of complex 18 to not only a very mild but also a space-demanding reducing agent.

Of particular importance is the in situ complexation of the strong and highly oxophilic dialkyl aluminum hydrides, for example, DIBAL [6, 11].

On treatment of the multifunctional indolo-quinolizidine 20 with a plain toluene solution of this reagent, one observes a very unselective and also unreliable reduction, leading to an unattractive mixture of compounds.

If, however, the toluene solution is pretreated with glycol dimethyl ether, the very selective and highly reproducible formation of hydroxyester 21 is noted [12].

The warming up of the hydride solution on addition of the diether indicates complex formation, to slow down the reactivity of the reducing reagent.

The high tendency for aluminum–oxygen interaction may also be responsible for the highly selective reduction of nitrile ester 22 with DIBAL in the absence of the diether at low temperature [13].

While the polarization of carbonyl groups and the Lewis base capacity of hydroxy groups offer a number of options for complexation, the situation is quite different with carbon–carbon double bonds.

Nevertheless, there are various possibilities to influence their reactivity along these lines too.

Neighboring hydroxy groups play a vital role in attracting and anchoring metal catalysts, which then deliver, for instance, hydrogen, into properly located double bounds.

This principle also operates very satisfactorily in oxidation reactions as the well-known and widely used Sharpless reaction clearly demonstrates.

For high chemo- and diastereoselectivity, the choice of the catalyst is of course essential and for hydrogenations very good results have been achieved with rhodium and iridium complexes [14].

With example 24, one should not overlook that the higher substituted double bond is hydrogenated and that the chemoselectivity of this process is accompanied by excellent diastereoselectivity. In addition, it turned out that the presence of isopropyl alcohol is mandatory for high chemoselectivity. In the absence of any complex-forming directing groups, there can be different chances for charge stabilization as an important prerequisite for selective attack at a carbon–carbon double bond.

The most simple approach could be the use of any type of Michael addition, employing strong acceptor groups such as esters, nitriles, or nitro groups.

Selective additions to these double bonds will certainly take place, but if the directing acceptor group is of no use in further operations, or maybe even absolutely unwanted, the subsequent removal of this moiety will be troublesome.

In contrast to this, trialkylsilyl groups can easily be removed and therefore offer themselves as charge stabilizer.

While alkyl substituted double bonds under normal conditions do not intervene in Grignard reactions, the trialkylsilane-substituted olefin 26 nicely forms a five-membered ring (27), generating a silicon–magnesium intermediate, which, representing an equivalent of a bis-anion, shows very high nucleophilicity.

The role of silyl groups as directing centers is gaining growing importance as this moiety serves as an excellent example to illustrate the general strategies for transition state manipulation [15].

On the one hand, these groups can take the role of an active volume, influencing the course of a reaction by charge stabilization (see 28), while on the other, space-demanding alkyl substituents, as in the TIPS-group (tris-isopropyl-silyl) (see 29), render them into passive volume, which means that they influence just by their sheer size.

The wide range of options to use silane groups of different reactivity for chemoselective transformations is nicely demonstrated by an example from the benzleukodienes (33) [16].

Having seen these impressive examples, we shall not be surprised by the silyl groups in the following chapters on regioselectivity and stereoselectivity.

Chemoselectivity poses particularly demanding problems if the same functional group is present at different positions of a molecule as in sugars or glycosides.

In this case, there may be options to rely on the sterical situation, especially if one can reversibly retreat to cyclic or bicyclic structures.

Very often, however, the assistance of protecting groups will have to be considered, at least as long as purely chemical transformations are employed.

There are quite encouraging signals, however, from various types of enzymatic reactions.

It is, unfortunately, absolutely impossible to discuss the progress and the future possibilities in this field in this chapter but we include at least one example to demonstrate the capacity of these tools [17].

It is hard to see that any type of conventional hydrolysis could compete with these results.

1.2 Regioselectivity

Regioselectivity is of particular importance with fundamental starting materials carrying functional groups that offer two reactive positions, such as olefins, acetylenes, epoxides, anhydrides, and imides. There are additionally the two enolate structures of ketones, as well as unsaturated carbonyl groups (1,2- vs 1,4-addition). In addition, there are a number of aromatic and heteroaromatic compounds posing various problems with regard to regioselective substitution.

With olefins, regioselectivity is governed by the Markownikov rule, but there are examples of anti-Markownikov additions, with hydroboration [18] being the most prominent one.

In case all these regulations leave deficits, one can still retreat to a few modifications of the double bond to solve the problem, as for instance, the epoxide, or the corresponding allylic or vinylic systems.

It has to be mentioned at this stage that triple bonds are posing very similar problems that are treated along the same lines.

It should be noted, however, that, in this case, hydroboration and analogous metal hydride additions generate the very useful vinyl anion equivalents 38, which nicely contribute to the synthetic methods for allylic systems [19].

While acetylenes add directly to aldehydes and ketones to give rise to the propargylic systems 39, which lend themselves for hydrogenation, the vinyl anions of type 38 lead directly to the corresponding allylic alcohols 40.

Up to this point, the regioselectivity can be taken for granted. This changes, however, when we turn to the palladium-catalyzed substitutions, which have been broadly investigated in this field, with particular emphasis on the corresponding carbonates [20].

Out of the many useful transformations published, we selected just two, to demonstrate that one has two options here, leading either to direct substitution 42 [21] or to the SN′-type products 41 [22].

While the SN′-process introduces a functional group at the olefinic 1,3-position (42), direct substitution can lead to a wide choice of allylic substituents. Both can influence the reactions of the remaining double bond in various ways.

In all these metal-catalyzed substitutions, the carbon framework operates as an allylic cation equivalent. Moreover, to steer the regioselectivity one relies mainly on leaving group properties and reaction conditions.

Very similar problems arise with allylic anions of type 43.

Regioselectivity will be particularly hard to achieve if there are only small differences in space demand and electronic properties between R′ and R″.

Under these circumstances, the electrophiles may not properly differentiate between α- and γ-positions.

Again, silicon comes to the rescue [23].

Owing to charge stabilization at the α-position anion 44 gives rise to the α-substituted homoallylic alcohol 46 while the bulky TIPS group directs the electrophile into the γ-position, generating the vinyl silicon compound 47.

The double bond in this product is again well prepared for highly regioselective transformations.

The corresponding epoxide 49, for instance, opens regioselectively at the β-position (β-effect of silicon!) and gives rise to aldehyde 48 via silicon migration [24].

It is noteworthy that in the course of this sequence both carbon atoms of the double bond become substituted in a highly selective and predictable manner.

In this case, we deal with the electronic effect of a neighboring silicon substituent, but simply properly chosen reaction conditions can efficiently determine the outcome of epoxide ring openings too.

As one would expect, the employment of an oxophilic Lewis acid leads to cation formation at the higher substituted carbon atom of the epoxide, while attack with a strong nucleophile takes place at the less substituted one.

Aluminum hydrides serve as perfect examples for this outcome. In the case of epoxide 50, the nucleophilic tetrahydrido anion attacks the α-carbon atom, leading to the tert-alcohol 52. In contrast to this, the Lewis acid DIBAL gives rise to the primary one (53) under reductive shift of the double bond (see 51) [6].

These observations lead to the general rule: Lewis acid reactions are governed by cation stability and pure nucleophilicity by steric effects.

The first case is demonstrated by the regioselective formation of the highly substituted amines 54 and 55 [25].

In this connection, one may consider the regioselective formation of the elimination product 56 in the presence of a Lewis acid lacking any nucleophile, proof of the mechanistic interpretation of Lewis acid–catalyzed epoxide splitting [26].

As far as purely nucleophilic ring opening is concerned, it is very rewarding to notice that all the well-described orbital overlap requirements that are very typical for the Walden inversion process are mandatory for the epoxide reactions too.

With rigid epoxide structures, this is nicely reflected in the well-established Fürst–Plattner rule, which demands diaxial orientation for the transition state (see 57) [27].

In accordance with the rule, the 3,4-epoxides of steroids or terpenes determine the regioselectivity of nucleophilic attack. Nucleophiles show up in the 3-position (58) with β-epoxides while the corresponding α-epoxides undergo nucleophilic ring opening at the 4-position (59).

This rule is extremely important for directing regioselectivity in rigid systems, and the high potential of this statement can be judged from regioselective opening to provide alcohol 57. This is obviously also governed by the Fürst–Plattner rule [28], in spite of the handicap of having to accommodate all substituents in axial orientation.

These results indicate that to exercise very reliable and predictive regioselective epoxide transformations one has to be well aware of the mechanistic details of the process.

This can nicely be demonstrated with the intramolecular ring opening reaction of epoxide 60 [29].

At first glance and ignoring stereochemistry, one is tempted to predict cyclopentane formation, but in this event the cyclobutane 61 is mainly formed.

Looking at the transition states, one is convinced that the carbon chain is simply too short to reach the trajectory for the first process (see dotted line), while the four-membered ring can easily be formed.

One has to realize that the stereochemical effect (axial substituents) as well as ring strain considerations are completely overruled by overlap necessities.

The final example in this series of epoxide reactions serves as a proof that the outcome of these reactions is independent of the nature of the anion involved and that an sp2-centered nucleophile follows exactly the same rules.

At very low temperature, the anion generated from vinyl sulfone 62 operates again in a highly regioselective manner, leading to dihydrofuran 63 [30].

Summarizing these results, we end up with two requirements: With Lewis acid catalysis it is the substitution pattern that counts, whereas orbital overlap is crucial for SN2-type reactions aiming at directed regioselectivity in epoxide ring fission. Compared to just these two parameters in the epoxide case, there is quite an arsenal of tools to manipulate enolate formation in ketones. Considering the high value of this functional group for bond-forming chemistry, one is not surprised to notice that a wide variety of options to manipulate enolate formation has been investigated.

They range from number, size, and electronic properties of α- or β-substituents (see 64) via ring size and rigidity to various derivatives of the carbonyl group such as oximes, alkylated oximes, and all types of hydrazone derivatives.

In addition, there is a multitude of variations from the side of the reagent. It starts with solvent, catalyst, and reaction temperature to continue with the size of the deprotonating species, the addition of countercations, and selected crown ethers.

Since these conditions may also control the transprotonation steps, we could also employ kinetic versus thermodynamic control.

Considering all this, very impressive results have been achieved already.

Deprotonation of ketones 65 and 67 with the bulky “Loba”-base, for instance, proceeds with very high regioselectivity (97%) to generate the less substituted enolates 66 and 68, quenched as silylethers [31].

Although this certainly meets our expectations – if not to a large extent – the deprotonation of hydroazulene-ketone 69 with lithiumtriphenylmethyl leading to mainly one enolsilylether is really remarkable [32].

After palladium oxidation, cyclopentenone 70 is obtained with at least 90% selectivity. The structural difference here amounts to just one methyl group in the γ-position. However, since the seven-membered ring shows quite some conformational mobility, simply counting heads could be misleading.

It is of course very tempting to combine sheer size of the proton acceptor with conditions of kinetic or thermodynamic control, as has been shown for α-methylcyclohexanone 71 [31b,c]

As these results show, proceeding in this manner is clearly of practical value, and Shea proved in a very detailed investigation that this strategy can be used quite efficiently for bridgehead substitution in the important bicyclic ketone 74 [33].

The anti-Bredt position of the 1,2-enolate, together with molecular mechanics calculations, indicates this to be the thermodynamically disfavored position.

Under kinetic control, however, it is formed with high selectivity. On methylation the bridgehead-substituted ketone 75 is obtained and on oxidation it gives rise to the bridgehead carbinol 77.

If the deprotonation is done under thermodynamic conditions, the 3-methyl derivative 76 results from the methylation process.

It is noteworthy that these results constitute a complete reversal of the regioselectivity in enolate formation and the 99° angle of the C1–H bond with the carbonyl group is a clearly convincing explanation of the high acidity.

As noticed with the epoxides, the intramolecular capture of enolates can be very helpful to solve regioselectivity problems too.

In the case of the bicyclic ketone 78, the plain thermal cyclization generates synthetically unattractive mixtures of the five- as well as the six-membered ring compounds 80 and 79.

If one starts with the separately prepared enolsilylether 81, mercury-catalyzed cyclization gives a high yield of 79, while the easy-to-make aldol 82 leads to 80 after a thermal retro-aldol process [34].

This example teaches that minor changes in the procedure can result in regioselective routes to both possible enols.

This means that to reach a special target one simply has to select the appropriate approach to the enolate needed.

Finally, this subtopic provides another generally very useful application of the active volume–passive volume principle.

Although ketone 83 at first glance appears to be a good candidate for selective alkylations or Michael additions, first experiments using the tris-methoxy compound 83b met with complete failure.

With methyl propiolate as the electrophile a very disappointing mixture of products was obtained. The picture fortunately changed completely with the monohydroxy compound 83a.

Michael addition with methyl propiolate led in this case directly to the tricyclic α-pyrone 84, which is a central intermediate in Eschenmoser's colchicine synthesis [35].

Obviously, the methyl group in 83b is not bulky enough to completely divert the electrophile from the benzylic position, thus giving rise to mixtures. To achieve complete shielding of this center, highly space-demanding groups such as pivalate, TIPS-ether, or maybe even the triphenylmethyl group will probably be necessary to create the appropriate passive volume.

In contrast to this, the free phenolic group in 83a presents itself as active volume, probably picking up the propiolate to form an enolether, which is then transferred to the benzylic position exclusively (see 85).

Under these circumstances, the generation of the “wrong” enolate will be of no consequence as long as enolate equilibration is guaranteed. The capture of the electrophile, in this case, represents a Michael addition to an acceptor-substituted acetylene, and this brings us to another subtopic in the carbonyl field that is bound to pose regioselectivity problems.

Conjugated triple bonds as well as double bonds can give rise to 1,2- or 1,4-addition products, and the picture will be even more disturbing if we should be confronted with the inverse Michael addition too.

In general, and as long as we deal with ionic additions, this process is governed by the hard–soft principle and in the case at hand, having potassium as the countercation together with a soft nucleophile, 1,4-addition can be taken for granted. An inspection of the general picture, however, reveals various possibilities to manipulate the outcome of these reactions (see 86).

The acetylenic amides of type 87 proved to be an excellent testing ground for this behavior [36].

Although the complexing amide is of assistance for the α-addition, the phenyl group diverts the nucleophile only to an extent of 10% to the α-position (see 89).

If, however, the tert-butyl group directs the approach, the α-addition product 91 is formed to the extent of 98%.

As different studies on the directing power of various passive volume groups have shown (see chapters 1 and 3 on stereochemistry), branched saturated substituents proved to be more space demanding than a phenyl group.

Things get a little more complicated with ambident nucleophiles and unsaturated ketones, as demonstrated with cyclopentenone 92 [37].

We notice here a quite strong solvent dependence, but although 93 formally appears a violation of the hard–soft principle, a zinc chelate such as 95 could easily explain this outcome.

The regioselectivity with ambident anions also shows strong dependence on Lewis acid catalysis, as was nicely demonstrated with enthiolat 96 [38].

Since all the reactions were run in tetrahydrofuran, the fact that three out of four possible products can be generated selectively is solely due to the Lewis acid present. While lithium as the countercation leads to the “normal” Michael adduct 98, titanium gives rise to the corresponding 1,2-adduct 97. Aluminum favors 1,4-addition again but employing the sulfur atom as the nucleophile (99).

The high reliability of 1,4-addition offers an excellent chance to capture the intermediate enolate anion in a completely regioselective and diastereoselective tandem process.

This highly flexible sequence opens an easy path to 2,3-disubstituted cyclopentanones (see 102), as has been impressively demonstrated in the total syntheses of the prostaglandins [39].

If a leaving group is placed properly in the cyclopentenone (see 103), the trans-protonation equilibrium 104a104b will regenerate a cyclopentenone (107) via β-elimination.

This can then be followed by a second 1,4-addition–enolate capture sequence to provide the trisubstituted cylopentanones 105 with complete regio- and stereoselectivity. As far as the reagents involved and their sequence of addition are concerned, one notices very high flexibility. Consequently, these reaction cascades offer easy-to-direct routes to a great number of differently substituted cyclopentanones, and if one additionally runs these cascades in an intramolecular manner, they grant access to numerous anellation products [40].

All efforts to preserve a charge or a radical in a molecule to show it around and use it for directed cascade-like reaction sequences at well-selected positions will be of great importance for the future development of an automatized and sustainable synthetic chemistry. One is therefore well advised to look for substituents or functional groups that can provide assistance in these endeavors. As far as charge stabilization is concerned, the trialkylsilyl group once again is very promising and has indeed been very helpful in Michael additions already.

The trimethylsilyl-substituted methylacrylate 108, for instance, first presents itself as a highly reactive Michael-acceptor, which can in the next step trigger a very efficient Peterson olefination to provide the acrylates 109 and 110 [41].

As the reaction scheme demonstrates, the nature of the donor and the electrophile can be exchanged, leading to a wide choice of substituents for substituted acrylates with predictable regioselectivity.

In a recent report on directed regiodivergent additions to a triple bond (see 111) the change of reagents is combined with an interesting change from ionic to radical initiated additions [42].

As can be expected, cuprate addition leads to β-attack, which is followed by the normal α-capture of the negative charge (see 112). Radical attack, on the contrary, takes place at the α-position to generate a resonance-stabilized intermediate, which is terminated at the β-position.

If at this last stage an aldehyde is employed, the lactone 113 is formed.

Discussing conjugated systems, we have until now dealt with unsaturated carbonyl groups exclusively. Actually, this is just one species in the large group of unsaturated acceptors, but the observations made here and the conclusions drawn will, by and large, be valid for the whole field.

So, we add just one example from the sulfur field, and the vinyl sulfoxide 114 presented was chosen for two reasons.

First, it may operate as a very normal Michael-acceptor. However, due to the deprotonation conditions it could also lead us into the area of vinyl anions, rendering the molecule into a nucleophile. This means that with a change of the reaction conditions, we should be able to switch into a completely different line of products [43].

Both processes, however, take place with perfect regioselectivity (see bonds printed in bold). The kinetically controlled formation of the heteroatom-stabilized α-vinyllithium intermediate is followed by a very quick nucleophilic attack at the carbonyl group, while thermodynamic control to generate the enolate gives rise to the normal conjugate addition to the acceptor-substituted double bond. Stabilized vinyl anions such as 117, 118, and 119 are, owing to their constitutional and configurational stability, quite interesting homoenolate intermediates useful for regioselective double bond transformations [44].

With the vinyl anion of butadiene the regioselectivity in the addition to carbonyl groups is strongly dictated by the countercation [45].

While the Grignard reagent 120a provides a roughly 80 : 20 mixture of 121 and 122, the corresponding lithium compound 120b leads to the α-adduct 122 with high selectivity.

As far as polyanions of type 123 are concerned, one can generally rely on the sequence of acidities of the corresponding protons, with the position of the lowest acidity exercising the highest nucleophilicity.

Regioselectivity in aromatic compounds is in principle dictated by substitution rules. With heteroaromatic rings, however, there is a strong influence of the involved heteroatom and its hybridization.

While pyrroles and analogous donor aromats provide substitution products easily and selectively, the acceptor aromats, for example, pyridine or quinoline, suffer from bad yields and low selectivity in standard electrophilic substitutions.

One way out of this dilemma is to bring along the substituents in question in a properly planned synthesis of the heteroaromat, but in some cases minor changes such as quarternization or N-oxide formation are of considerable help already.

With quarternized compounds, we switch from electrophilic to nucleophilic attack and with N-oxides we improve the donor capacity of the aromatic ring.

While pyridine itself leads to 3-nitropyridine under quite drastic conditions only, the 4-nitropyridine N-oxide is formed smoothly and in high yield.

With benzene and its derivatives, the various types of coupling reactions developed in the last decades open highly reliable routes to a wealth of substitution patterns [46–49]. In most cases, however, the corresponding halogen compounds are needed as starting materials, which does not change the substitution problem. Still, the scope is so wide that it cannot be treated in detail here.

One particular and synthetically demanding problem that requires intelligent solutions represents the generation of ensembles of ortho substituents.

This task unfolds again as a very typical case of passive volume–active volume principle.

In classical substitution reactions, ortho substitution suffers from steric hindrance. The sheer bulkiness of the substituents present blocks the substitution process.

Should the substituent present itself as active volume, however, assisting in the interaction with the approaching reagent will of course completely change the situation.

A large number of ortho-directing substituents have been developed for metalation reactions and 124–126 represent a few typical examples [50].

Since aromatic fluoro compounds are gaining rising importance in the field of biologically active structures, we are adding at least a few reactions to demonstrate the directing capacity of the fluoro substituent. Preferential deprotonation with sec-butyllithium can be achieved in both o-positions of aromatic fluoro compounds.

In the case of the protected p-fluorphenol 127, this leads to 128 and the subsequent quenching with alkylborate gives rise to 130, which can be used for coupling reactions as well as for substituent exchange.

The standard oxidation procedure, for instance, provides phenol 129.

After protection, 129 undergoes a second deprotonation and capture of the corresponding o-lithium intermediate with dimethylformamide, and finally generates o-fluoraldehyde 131. Exercising exactly the same sequence twice with the protected p-fluorbenzylic alcohol 132 takes us to the tris-silyl compound 134, which, after deprotection, oxidation, and treatment with benzylic bromide provides the p-fluoraldehyde 133 [51].

From the examples offered, one could draw the conclusion that regioselectivity is in the sole responsibility of neighboring substituents.

That this picture is too simple emerges clearly from the Grignard reactions of anhydride 135.

The results demonstrate very clearly the decisive role of the solvent.

In diethyl ether, the complexed reagent attacks quite slowly and leads via a late, product-orientated transition state to 136 with low selectivity, while the bulky and highly reactive TMEDA (tetramethylethylenediamine) reagent proceeds via an early transition state at the less hindered m-carbonyl, giving rise to 137 [52].

With anhydride 135, we demonstrated the influence of the solvent on the reactivity of a metallo-organic reagent, resulting in the attack at special positions of the aromatic ring. With quinole acetate 138, the correct choice of the solvent even opens an opportunity to direct an alkyl substituent either into a ring position or into a benzylic carbon atom [53].

It is probable that in dichloromethane the quinole acetate is attacked in the allylic position, similar to the well-established Sakurai process, while in the more polar acetonitrile the acetate proceeds to the elimination product 142, which is then attacked by the nucleophile at the least hindered carbon atom.

In this section, we have described some important strategies to influence and direct the chemical transformations at ambident positions in order to arrive at either one or the other possible reaction product regioselectively.

In the following sections, this will be extended to stereoselectivity and enantioselectivity.

1.3 Stereoselectivity

The reliability and the fundamental nature of the stereoelectronic principle contribute heavily to the importance of this rule as the backbone of diastereoselectivity.

Since it provides the bridge between sp2- and sp3-centers and since very useful and quite flexible routes lead from acetylenes (sp) to even polysubstituted olefins (sp2) in a stereoselective manner, the sp → sp2 → sp3 sequence is of utmost importance for the whole field of stereochemistry.

Together with this sequence, the stereoelectronic principle represents the key for devising and understanding stereoselective transformations.

Depending on the electronic nature of the double bond, it paves the path to a very broad range of sp3-centers in a highly selective manner. The high synthetic flexibility observed is due to the nature and the properties of the electrophile as well as of the nucleophile. Its intervention via an early 143 or late 145 transition state will determine the stereochemical outcome of the process.

This calls for stereoselective, flexible, very reliable, and if possible simple techniques to prepare di-, tri-, and tetrasubstituted double bonds.

Although the Wittig reaction may, at first glance, appear a good choice for the preparation of olefins, one runs into difficulties very quickly when it comes to the synthesis of higher substituted olefins.

The outlook is much better with the Peterson reaction; here, at least trisubstituted double bonds with E- as well as with Z-configuration will result if the syn-elimination mode 147 or the corresponding anti version 148 is employed.

Very high chemical as well as configurational flexibility was secured from addition reactions to triple bonds, particularly when intramolecular steps promised regioselectivity.

An early, groundbreaking and easy to direct process was found in connection with aluminum hydride reductions of acetylenic esters 149 [54]

Since 151 appears a convincing intermediate in the addition process, the reaction was quenched with iodine to generate vinyl iodide 153, which, on treatment with a cuprate, gave rise to the well-defined trisubstituted olefin 152.

To obtain the alternate configuration 155 one simply has to exchange the substituents at the triple bond and the cuprate complex (see 154).

In the presence of a Lewis acid catalyst (AlCl3) the regioselectivity of iodine addition is inverted, leading to the substitution pattern 157.

Exchange of substituents as described above widens the scope of the structures that can be addressed with this method.

The high degree of flexibility observed in triple bond additions is certainly the reason for the numerous contributions to this field.

The investigations were extended from acetylenic esters to silyl compounds, sulfones, borates, as well as alkyl- and arylacetylenes [55].

The story would still be incomplete, however, if one would not add the many options that additionally arise from general metal hydride additions as indicated in 158.

Although all this could lead to the conclusion that acetylenes are great performers and the starting material of choice for directed selectivity in olefin synthesis, one should not forget the enol derivatives of ketones and also their hydrazones as useful starting materials [56].

With a β-ketoester such as 159 there are very good chances to easily secure the desired double bond configuration, and the cuprate nucleophile offers additionally a very broad choice of possible substituents in 162 and 163.

Comparable results have been reported for enol tosylates [57], such as 164, and the hydrazone 166 served very well for the preparation of the aryl-substituted cyclic olefin 167 [58].

It is not the intention here to cover this field comprehensively but rather to show that there is a number of quite different and easy to direct stereoselective routes into that area.

For much more detailed information, we recommend Negishi's very broad and well-organized review article [54]a.

With the properly substituted olefins at hand, we now have to address the numerous stereospecific reactions that can be employed in sp2 → sp3 transformations.

Next to the standard electrophilic additions, the formation of haloethers and halolactones is of particular synthetic value since epoxidations and diol formation can also be run with high face selectivity; these reactions offer very short and predictable routes from simple olefins to well-defined diastereomers and enantiomers.

Generally, the directed selectivity in epoxidations is easily accomplished with the assistance of a neighboring hydroxyl group. Accepting hydrogen bonding as one important phenomenon in this behavior, one is not surprised that other groups showing hydrogen bonding capacities too can also fulfill this job, as demonstrated with olefin 168 [59].

The ether group in 168a represents a passive volume and directs the oxidant to the α-side (see 169), while the amide group in 168b operates as active volume that picks up and delivers the reagent from the β-side.

If in intramolecular versions, comparable to iodolactonization, the nucleophile is represented by a carbon–carbon double bond or triple bond, the stereospecific and stereoselective generation of carbon bonds is at reach (see 171 and 172).

These π-cyclizations are, on the one hand, impressive examples of directed selectivity, since the configurations of the double bonds involved dictate the configuration of reaction products, but they can, on the other hand, also be run like a zipper though a polyolefin chain to generate pure stereoisomers of polycylic compounds, as for instance, in the squalene monoepoxide cyclization that represents the backbone of the biogenetic pathway to steroids [60].

Manipulations of constitutions and configurations can again be accomplished by exchanging double bond substituents and configurations (see 173/174).

Although we have so far seen quite a number of transformations that lead to synthetically useful compounds, we, as far as the mechanism goes, have not yet left the subject of electrophilic attack at π-systems.

Of course, there are many alternatives to transfer sp2 configurations into well-defined sp3-centers and certainly hydrogenation is a very simple and obvious one [14b, 61].

Keeping in mind that the substrate will have to interact with the surface of the catalyst that transfers the hydrogen atoms, one may safely predict that the configuration and conformation of the starting material will have a strong influence on the steric course of the hydrogenation.

Owing to the conformational flexibility of the seven-membered ring, the hydroazulenes 175 and 176 serve as convincing examples for this behavior [62].

The β-orientated γ-lactone in 175 clearly shields the upper side of the hydroazulene, directing the hydrogen into the α-orientation (see 176). If this particular ring is opened, one gains free access to the β-side, leading to 178.

Needless to say that hydrogenations in general and also other processes employing heterogeneous catalysis show a high dependence on substrate configuration and conformation too.

Again, we have to recommend review articles here for further information [14b, 61].

Out of the many additions possible, the wide field of cycloadditions certainly merits special scrutiny.

As in 2π-4π cycloadditions, two new bonds are formed in a highly stereoselective manner, the transition state demands are very high.

They are even more strict in intramolecular cycloadditions that they not only nicely solve regioselectivity problems but are also highly recommended for the introduction of quaternary carbon atoms [63, 64].

This works even when these centers are in close neighborhood as is demonstrated with ketone 182 [65].

The generation of two neighboring quarternary centers is, however, only one important aspect of intramolecularization; one should additionally realize that there is an impressive potential to manipulate the outcome of the process.

By changing the length of the carbon chain and the configuration of double bonds, this reaction can be directed into a manifold of various constitutions and configurations in the realm of polycyclic compounds.

It has to be remembered again that owing to their highly organized transition states these cycloadditions are completely stereospecific, which means that the configuration of the reaction products is determined by the configuration of the double bonds involved.

As both double bond configurations are easily available, one can switch from one diastereomer to the other just by employing the corresponding starting material, as was nicely shown with maleic and fumaric acid [66].

The trans-configuration of lactone 187 proves that in this case this particular carbonyl group occupies the exo-position in the transition state. This kind of violation of the endo-rule is, however, not unusual in intramolecular cycloadditions.

This observation indicates very clearly that if strain is imposed on the transition state conformation, the molecule may deviate into another arrangement, thus opening the path to an alternate configuration for the cycloaddition product.

This grants options for directed selectivity by manipulating transition state conformations with the help of space-demanding control groups. These will enlarge the rotation barriers favoring just one reaction path.

Again the trialkylsilyl group served very well in this endeavor.

While diester 188a on heating gives rise to a mixture of stereoisomers, the corresponding vinylsilane 188b provided the single product 189 in 85% yield [67].

Models indicate very clearly that other transition state conformations but 188 are destabilized by steric interactions with substituent X. 188 only does not suffer from steric hindrance.

For the same reason, decalin 191 was the sole (89%) cycloaddition product from trimethylsilane diene 190b.

It is very important for the application of this principle that the control element can be easily introduced into the carbon framework and that smooth removal would be guaranteed.

As even these very few examples show, intermolecular and intramolecular 2π-4π cycloadditions are the reactions of choice for stereoselective preparation of cyclic and polycyclic compounds.

It should, however, not be forgotten that next to the construction of carbon–carbon bonds these cyloadditions can also be used for the directed functionalization of selected carbon atoms.

With the nitroso group operating as a dienophile for instance, a wide range of precursors for α-amino alcohols becomes easily available.

With simple dienes there is a regioselectivity problem, but electronic orientation or intramolecularization is again very helpful as examples 192 and 193 indicate respectively [68].

Since the N–O bond can easily be broken by reduction, 192 presents itself as a precursor for substituted α-amino acids, while 193 can lead to substituted δ-lactams [69].

A very similar development took place with 1,3-dipolar cycloadditions.

As with Diels–Alder reaction, the early phase of this cycloaddition was completely dominated by the intermolecular construction of heterocycles (see 194 and 195).

It took quite a few years for the first examples of the intramolecular version to be published and it right away proved also to be an excellent tool for the directed introduction of functionality into alicyclic as well as heterocyclic frameworks [70].

Ring formation accompanied by stereoselective functionalization takes place in a very simple way with pyrrole aldehyde 196 [71], while in the case of the homoallylic alcohol 198 a straight carbon chain is stereoselectively transformed into the trisubstituted pyrrolidine derivative 199 [72].

The vinyl-nitro compound 200 serves as an excellent example to demonstrate the remarkable flexibility and the very high synthetic potential for forming heterocyclic compounds in a highly predictable manner [73].

It needs only a little bit of imagination to see how the carbon chains as well as the double bond determine the structure of the heterocyclic compound.

One should not overlook the great advantage of generating the 1,3-dipole en route to the cycloaddition step.

Another quite unique way of this in situ formation of the dipole was described in connection with the total synthesis of the spiro-piperidine system of histrionicotoxin [74].