Science of Synthesis

DOI: 10.1055/sos-SD-008-00001

Majewski, M.; Snieckus, V., Science of Synthesis, (2006) 8, 1.

In this Volume 8 of Science of Synthesis, 41 authors (M. Barbero, L. Brandsma, D. Caine, J. V. Comasseto, R. L. O. R. Cunha, R. K. Dieter, T. Durst, J. Eames, R. W. Friesen, R. E. Gawley, J. R. Green, G. W. Gribble, F. Hasanayn, A. Jończyk, E. Juaristi, R. M. Kellogg, M. Khodaei, R. Klein, A. Kowalkowska, A. P. Krapcho, R. Łaźny, S. MacNeil, M. Majewski, R. Melgar-Fernández, C. Metallinos, A. Mordini, O. Muñoz-Muñiz, C. Nájera, S. O'Connor, N. Ono, J. N. Reed, C. C. Silveira, V. Snieckus, A. Streitwieser, C. F. Sturino, M. Valacchi, P. Venturello, U. Wietelmann, L. Xie, M. Yus, and J. W. Zwikker) describe formation and use of a variety of compounds of the group 1 elements (Li, Na, K, Rb, and Cs) in organic synthesis. Lithium compounds are described in Volume 8a, while compounds of the other group 1 elements are covered in Volume 8b. This volume concludes the eight-part series in Category 1 (Organometallics) following in the elegant footsteps of the alkali metal coverage in Houben–Weyl, Vols. 13/1 (1970) and E 19d (1993). For the presentation of methods in chemical synthesis, the major difference between Houben–Weyl and Science of Synthesis is the comprehensive and exhaustive nature of the former and the selective, critically evaluated, and most useful and reliable methods presented in the latter.

A major challenge in planning the volume concerned the ubiquitous nature of alkali metal compounds in organic synthesis. Ranging from very simple reagents, such as sodium hydroxide, used to generate organometallic intermediates or as sources of cations designed to involve themselves in coordination (e.g., lithium and cesium salts) to “true organometallics” such as alkyllithium compounds, alkali metal derivatives are encountered in a plethora of experimental procedures. As may be appreciated, due largely to their simplicity, many useful alkali metal reagents are difficult to search using standard databases. Therefore, we decided near the beginning that Volume 8 shall depart from the pure “organometallics” theme and that some simple compounds used either for generation of polar organometallics, in situ or otherwise, will also be covered. Similarly, the alkali metal enolates are covered although arguably they are not organometallic species. This choice of coverage resulted in a rather sizeable two-part volume that nonetheless is rather selective in nature. As a consequence of a multiauthor work of this nature, overlap of content is inevitable and, we would argue, desirable for the completeness of a given topic, in order to provide individual flavor and perspective by a given author, and therefore to offer a different appreciation by the reader. For example, the generation of vinyllithium compounds may be viewed from the perspectives of applications of lithium metal (Section ‌8.1.1‌), of deprotonation of alkenes (Section ‌8.1.8‌), and of synthesis of α-lithio vinyl ethers (Section ‌8.1.27‌). Thus, by way of illustration, the chemist in search of selecting the best method for alkylation of a C—H acid is advised to consult the sections on the corresponding lithium, sodium, and potassium derivatives, and also the relevant sections on the most likely bases for the deprotonation step (inter alia, sections covering lithium amides, lithium and sodium hydride, sodium hydroxide, and phase-transfer catalysis.)

A number of major and distinct differences of Volume 8 compared to most Science of Synthesis volumes are readily identified. In particular, a product subclass is defined as follows:

(a) A reagent that acts upon organic molecules to generate alkali metal species, e.g. lithium hydride for the synthesis of lithium carboxylates (Section ‌8.1.2.1‌).

(b) A carbanionic, normally unstable and fleeting, organometallic intermediate and not an isolated product, e.g. sodium acetylide (Section ‌8.2.8‌). There are exceptions of alkali metal organics which may be isolated, characterized, and stored for certain periods of time.

(c) An inorganic reagent based on a group 1 metal that is useful in synthesis but has a tenuous connection to polar organometallics (e.g., sodium halides, Section ‌8.2.3‌).

With one exception, the compounds of each metal constitute a separate product class (Section ‌8.1‌: lithium, Section ‌8.2‌: sodium, Section ‌8.3‌: potassium, Section ‌8.4‌: rubidium and cesium). Within each product class are given the methods for particular members of the subclasses. The various synthetic intermediates of the given alkali metal are presented in a logical order in which inorganic reagents (e.g., sodium hydride, sodium cyanate) precede organometallic species (e.g., sodium cyclopentadienide) as outlined in the table of contents.

Therefore, the center of attention is the preparation of the intermediate species as emphasized in the discussion of each category, in the tables, and especially in the delineation of experimental procedures in the methods and variations sections. After all, a stated aim of Science of Synthesis, in both printed and, for the computer-reliant chemists of the future, electronic forms is to be the first-call source for synthetic practitioners.

Information on the scope and limitations of reactions of the intermediate species is abundantly but selectively provided in the discussion, tables, and experimental procedures. In the procedures, either the generation of the organometallic species only is given and its reactions are referred to in the text or the formation of the species is outlined, as is its treatment with a selected reagent, the workup of a reaction mixture, and the isolation of a product. Generally, physical, but not spectroscopic, properties are given for the products. Mechanistic discussion is minimal but references to original sources are provided.

Applications of product subclass sections follow the order of methods and, where relevant, variations of methods. Some product subclasses concerning reagents and generated organometallic species lend themselves well to presentation of applications for the preparation of specific classes of organic compounds (e.g., Section ‌8.1.27‌) whereas others are sufficiently defined in terms of their synthetic utility within the synthesis of the product subclass sections (e.g., Section ‌8.1.23‌) and therefore require no applications sections.

Perusal of this volume will clearly demonstrate the dominance of organolithium reagents and intermediates in organic synthesis compared to the other alkali metals. As a consequence, and not surprisingly, the previous extensive studies on structure, acidity (pKa), reactivity, mechanism, and theory of the “Polare Organometalle”,[‌1a‌] originally concentrating on organosodium and organopotassium compounds, have been superseded by extensive research on organolithium compounds. Thus, Houben–Weyl, Vol. E 19d, the theoretical–experimental treatment by Sapse and Schleyer,[‌2a‌] the pedagogic and colossal review by Schlosser,[‌3a‌] and the monumental volumes of Rappoport and Marek[‌4a‌] are rich sources for increasing our understanding of these highly basic and nucleophilic organometallics and thereby lead to greater ability to devise new base/ligand combinations, improve established reactions, and predict new ones. “For those who seek to discover new reactions, the most insightful lessons come from trying to trace important reactivity principles back to their origins.”[‌5a‌]

Unfortunately, structural information has been by and large insufficiently assimilated by synthetic chemists perhaps due to the lack of easily appreciated and widely encompassing predictive rules which would allow direct experimental tests. Analogous to the area of synthetic radical chemistry, which was illuminated by preceding physical organic studies, the pioneering efforts of Bauer,[‌6a‌] Beak,[‌7a‌] Boche,[‌8a‌] Collum,[‌9a‌,‌10a‌] Fraenkel,[‌11a‌] Hoppe,[‌12a‌] Lappert,[‌13a‌] Meyers,[‌14a‌] Schleyer,[‌15a‌] Seebach,[‌16a‌] Streitwieser,[‌15a‌] and Williard[‌17a‌] have significantly advanced our understanding of organolithium compounds with the consequent impact on synthesis.

In the area of polar organometallics, in spite of the lack of detailed knowledge of the species involved, their state of aggregation, and the mechanism of reaction with organic molecules, organolithium chemistry continues its positive march in synthesis, providing new valuable reactions, including increasing numbers which are applicable on very large scale. For example, the Merck–DuPont Merck synthesis of the angiotensin II inhibitor Losartan is a significant demonstration.[‌18a‌–‌20a‌] Similarly, the synthetic chemistry of enolates continues to grow,[‌21a‌] and although great advances have been made in unraveling enolate structure, the pithy comment made by Seebach in 1988, pointing out that considering our poor understanding of the nature of the species involved, it is remarkable how many reactions can actually be run with predictable results, continues to be relevant.[‌22a‌] Undoubtedly as a result of the strong basicity and high coordination and aggregation properties of organolithium compounds, their use in enantioselective synthesis has witnessed the most spectacular advance in this field since the mid-1990s.[‌23a‌,‌24a‌] For the future, in view of the expense of lithium, breakthroughs in catalytic reactions of organolithiums may be anticipated.

The sections dealing with sodium and potassium compounds give unique overviews of the use of inorganic salts of these alkali metals in organic synthesis. The wide use of sodium hydroxide and potassium hydroxide, especially in phase-transfer catalysis, and the corresponding sodium and potassium alkoxides for deprotonation of C—H acids having moderate pKa values reflects the availability and low cost of these materials. The use of sodium and potassium amides in organic synthesis has nearly disappeared from the literature since the 1970s, these strong bases being superseded by lithium dialkylamides and alkyllithiums. Organosodium and organopotassium compounds derived from sp2 (aromatic, vinyl) and sp3 organic compounds have not, in spite of lower cost, enjoyed general applicability owing to the difficult handling of these unstable and pyrophoric substances. In small-scale chemistry, the highly powerful Lohmann–Schlosser base combinations (R1Li–R2OK) are widely used.

From the rubidium/cesium product class, the special characteristics of cesium salts in certain organic reactions (such as the “cesium effect” in macrocyclization and cesium bases in Suzuki–Miyaura cross coupling) are recognized by synthetic chemists.

Finally, perhaps an unnecessary statement is that this volume, as all others in the Science of Synthesis series, will make its profound impact via the periodic electronic updates.

Volume 8b is the second part of the review of alkali metal chemistry from the perspective of organic synthesis and deals with sodium, potassium, cesium, and rubidium, and their selected derivatives. Organolithium compounds and other lithium species are described in Science of Synthesis, Volume 8a. Taken together, Volumes 8a and 8b conclude the eight-part series in Category 1 (Organometallics) following in the elegant footsteps of the alkali metal coverage in Houben–Weyl, Vols. 13/1 (1970) and E 19d (1993). For the presentation of methods in chemical synthesis, the major difference between Houben–Weyl and Science of Synthesis is the comprehensive and exhaustive nature of the former and the selective, critically evaluated, and most useful and reliable methods presentation of the latter.

As has been already pointed out in the introduction to Volume 8a, a major challenge in planning the volume concerned the ubiquitous nature of alkali metal compounds in organic synthesis. Broadly speaking, Volume 8 covers the alkali metals in diverse and arguably unusual categories: (a) their elemental form; (b) a number of simple derivatives such as hydrides, hydroxides, and halides; (c) “true organometallic compounds” bearing metal—carbon bonds such as alkylmetals; and (d) metal enolates. Hence, Volume 8 departs to some extent from the purely organometallics focus of the eight-part series. From the above defined categories, the simple alkali metal derivatives are most difficult in terms of information retrieval for a review because they do not appear as keywords, are too common to be abstracted, and represent a daunting number of reagents and compounds. “Simple” derivatives (category b) are encountered in a plethora of experimental procedures and are undoubtedly used in the most diverse ways. They are also among the oldest reagents of organic chemistry; sodium hydroxide, for example, was in use during the Berzelius/Woehler era.[‌1b‌]

Chemical properties and the consequent synthetic applications of derivatives of alkali metals overlap to a large degree, i.e. the behavior of sodium hydride and potassium hydride can be expected to be similar. As a result, and also owing to the nature of multiple authorship, overlap of content between chapters is inevitable and, we would argue, desirable for the completeness of a given topic, in order to provide individual flavor and perspective by a given author, and therefore to afford a different appreciation by the reader. However, a cautionary note regarding the perceived overlap of expected chemistry is in order: selecting sodium hydroxide instead of potassium hydroxide or, similarly, a sodium enolate instead of the analogous lithium enolate for a given reaction might be a nontrivial issue. Some comments on differences in reactivity depending on the metal involved are provided in the overview of Volume 8b below.

Clearly, the chemist consulting Science of Synthesis is expected to be primarily interested in finding methods and procedures for the synthesis of defined targets. For this purpose, the broader the perspective, the better and therefore, as a reference source, Volumes 8a and 8b should be consulted side by side. To illustrate, the chemist in search of selecting the best method for alkylation of a C—H acid is advised to consult the sections on the lithium, sodium, and potassium species corresponding to the C—H acid, as well as the relevant sections on the most likely bases to be used for the deprotonation step (inter alia, lithium amides, lithium and sodium hydride, sodium hydroxide, and phase-transfer catalysis). In addition, organomagnesium and organozinc compounds should be considered as they are a part of the polar organometallic landscape and, in some reactions, such as nucleophilic addition to carbonyls, are synthetically equivalent to alkyllithium, alkylsodium, or alkylpotassium reagents. The reader is referred to Science of Synthesis, Vol. 7 [Compounds of Groups 13 and 2 (Al, Ga, In, Tl, Be … Ba) (Section ‌7.6‌)] and Vol. 3 [Compounds of Groups 12 and 11 (Zn, Cd, Hg, Cu, Ag, Au) (Section ‌3.1‌)] for discussion of these compounds.

A number of major and distinct differences of Volume 8 compared to most Science of Synthesis volumes are readily identified. In particular, a product subclass is defined as follows:

(a) A reagent that acts upon organic molecules to generate alkali metal species, e.g. sodium hydride for the synthesis of sodium alkoxides (Section ‌8.2.2‌).

(b) A carbanionic, normally unstable and fleeting, organometallic intermediate and not an isolated product, e.g. vinylsodium (Section ‌8.2.7‌). There are exceptions of alkali metal organics which may be isolated, characterized, and stored for certain periods of time.

(c) An inorganic reagent based on a group 1 metal that is useful in synthesis but has a tenuous connection to polar organometallics (e.g., sodium halides, Section ‌8.2.3‌).

With one exception, each metal constitutes a separate product class: [Section ‌8.1‌: lithium (see Vol. 8a); Section ‌8.2‌ sodium, Section ‌8.3‌: potassium, Section ‌8.4‌: rubidium and cesium]. Within each Product Subclass are given the methods of preparation of particular members of the subclass. The various synthetic intermediates of the given alkali metal are presented in a logical order in which inorganic reagents (e.g., NaH, NaCN) precede organometallic species (e.g., sodium cyclopentadienide) as outlined in the table of contents.

Therefore, the center of attention is always the preparation of an intermediate species for a given reaction as emphasized in the discussion of each category, in the tables, and especially in the delineation of experimental procedures in the methods and variations sections. After all, a stated aim of Science of Synthesis, in both printed and, for the computer-reliant chemists of the future, electronic form is to be the first-call source for synthetic practitioners.

Information on the scope and limitations of reactions of intermediate species is abundantly but selectively provided in the discussion, tables, and experimental procedures. In the procedures, either the generation of the organometallic species only is given and its reactions are referred to in the text, or a complete procedure in which the generation of a species, its treatment with a selected reagent, the workup of the reaction mixture, and the isolation of a product are fully elaborated. Generally, physical, but not spectroscopic, properties are given for the products. Mechanistic discussion is minimal but references to original sources are provided.

Application of product subclass sections follow the order of methods and, where relevant, variations of methods. Some product subclasses concerning reagents and generated organometallic species lend themselves well to presentation of applications for the preparation of specific classes of organic compounds whereas others are sufficiently defined in terms of their synthetic utility within the synthesis of the product subclass sections and therefore require no separate applications sections; this is often the case for reactive intermediates.

Volume Perspective

To provide the reader with a perspective on Volume 8b, the major features are summarized below. Details can be found in the relevant sections.

The discussion in each subclass begins by consideration of the metal in the elemental state (Sections ‌8.2.1‌, ‌8.3.1‌, and 8.4.1). The chemistry of group 1 metals is dominated by their low ionization energy, a feature that is responsible for their major use in reduction reactions (‌Scheme 1‌).

‌Scheme 1‌ Typical Reductive Processes with Elemental Alkali Metals

All of the alkali metals in dry ammonia or in an ether solvent establish an equilibrium of a solvated electron and the corresponding solvated cation with the electron ready for transfer to a variety of functional groups. In general, the heavier the metal, the greater the reducing power of the solution. Even though a large number of other reduction protocols such as those based on catalytic hydrogenation or metal hydrides are now available, methods that rely on alkali metals are still eminently useful, being often less expensive and frequently proceeding with high chemo- and stereoselectivity. Perhaps most popular in this class is reduction involving sodium metal in liquid ammonia, which is applicable to a great variety of functional groups, and includes dearomatization (the “sodium” Birch reduction) and reductive (pinacol and McMurry) coupling procedures (‌Scheme 2‌).[‌2b‌–‌6b‌]

‌Scheme 2‌ Examples of Reductions with Alkali Metals

A classical expression of stereoselectivity of this procedure, which is controlled by thermodynamics, is encountered in reduction of alkynes to E-alkenes, and the observations on the differences between lithium and sodium in this reaction now constitute a timeless contribution to chemistry (‌Scheme 3‌).[‌7b‌] As evident from the experimental procedures in the appropriate sections, some effort has been expended toward developing special forms of alkali metals with altered reactivity such as “sodium sand”, “micronized sodium”, potassium intercalated in graphite or absorbed on alumina, and derivatives such as superactive alkali metal hydrides.[‌8b‌] Clearly, selection of the appropriate reagent is important. Common reductive processes encompass reduction of aldehydes and ketones (including reductive coupling protocols), reduction of alkenes and alkynes, and also reductive cleavage of C—O, C—N, carbon—halogen, and C—S bonds.

‌Scheme 3‌ Reduction of Alkynes with Lithium and Sodium[‌7b‌]

Sodium and potassium hydrides are used primarily as bases to generate alkoxides, enolates and, if the term may be accepted,[‌9b‌] various “carbanions”. Carbanions can be used further as bases better suited to the specific system, e.g. the dimsyl anion [MeS(O)CH2−]. Use of sodium hydride and potassium hydride in the generation of borohydride reagents deserves special mention {see Science of Synthesis, Vol. 6 [Boron Compounds (Section ‌6.1.2‌)]}. Typically, the use of an alkali metal hydride results in the formation of a polar intermediate and this is followed by a reaction with an electrophile (e.g., alkylation of carbonyl compounds, addition to a carbon—heteroatom double bond, etc.) or with a proton donor (e.g., the Nef reaction), or, less frequently, by a rearrangement or fragmentation (e.g., fragmentation of homoallylic alkoxides or oxy-Cope and other rearrangements, ‌Scheme 4‌).[‌10b‌] Virtually every nucleophile–electrophile reaction has an intramolecular variant, e.g. the Dieckmann condensation. Often the reacting system is complex and the alkali metal hydride is used in one of several steps or as a component in a mixed reagent, e.g. metal-containing complexes used as reducing agents (Section ‌8.2.2‌) and potassium hydride/oxygen/crown ether systems used in oxidation (Section ‌8.3.2‌). An example of a multistep process that involves the hydride only in one step is the palladium(0)-catalyzed deconjugative allylation of malonates,[‌11b‌] which can be viewed from the perspective of a reaction involving sodium hydride (this point of view highlights the choice of the base used in generation of the nucleophile), as one of many reactions of metal enolates (focusing on the structure of the enolate), or by considering it as a part of the more general landscape of organopalladium chemistry (vide infra).

‌Scheme 4‌ Tandem [2,3]-Wittig–Anionic Oxy-Cope Rearrangement[‌10b‌]

While the selection of sodium hydride over potassium hydride might be dictated by safety factors (KH is much more reactive and more of a safety hazard), there may be better reasons for selecting one of these two seemingly very similar reagents. For example, dimsylpotassium (generated with KH) is superior to dimsylsodium (generated with NaH) in Hakomori methylation of carbohydrates.[‌12b‌] The basicities of sodium hydride and potassium hydride appear to be harder to quantify than, for example, those of alkoxides and may be somewhat dependent on the means of preparation of the hydride.[‌13b‌] Major applications of alkali metal hydrides are summarized in ‌Scheme 5‌. Details, examples, and specialized applications of hydrides are found in Sections ‌8.2.2‌ and ‌8.3.2‌. Sodium and potassium hydrides are also used in generation of metal enolates and, for these, the reagent might have a bearing on the process (vide infra).

‌Scheme 5‌ Typical Uses of Alkali Metal Hydrides

Alkali metal hydroxides, alkoxides, carbonates, and acetates comprise a large group of reagents that can be classified as metal–oxygen derivatives. These reagents are primarily used as bases and a number of special protocols have been developed, of which phase-transfer catalysis (PTC) is the most important. Applications of these compounds as bases for deprotonation (generation of nucleophilic anions) are rather analogous to reactions of sodium hydride and potassium hydride and follow the schematic presented in ‌Scheme 5‌. A number of well-known reactions, such as condensations involving carbonyl compounds (aldol, Claisen, Dieckmann, Darzens, Robinson annulation, inter alia), C—C bond-forming reactions via phosphorus, sulfur, or nitrogen ylide intermediates (e.g., the Wittig, Horner–Wadsworth–Emmons, Ramberg–Bäcklund reactions), addition of carbanions or other anions to multiple C—C bonds (Michael reaction and its variants), and elimination reactions (e.g., dehydrohalogenation) rely on the use of hydroxides and alkoxides as bases. Unlike the hydrides, the alkali metal–oxygen reagents are also useful as nucleophiles and participate in such textbook reactions as nucleophilic aliphatic substitution (e.g., Williamson ether synthesis, cleavage of epoxides), vinylic nucleophilic substitution, addition to carbonyl groups (e.g., ester hydrolysis, transesterification, ring opening), nucleophilic aromatic substitution, and others.

Two special subgroups of reactions involving alkoxides and hydroxides are cross-coupling processes, where alkoxides can act as activators and have a great influence on the reaction outcome,[‌14b‌] and cleavage of silyl enol ethers, where alkoxides can be used to generate the alkali metal enolate from the silyl derivative.[‌15b‌] Typical reactions of alkoxides and hydroxides acting as nucleophiles are shown in a generalized format in ‌Scheme 6‌, which is meant to be illustrative rather than exhaustive.

‌Scheme 6‌ Typical Reactions of Alkoxides and Hydroxides as Nucleophiles, and Use in Cross-Coupling and Desilylation Reactions

While the popularity of sodium ethoxide and potassium tert-butoxide may be a matter of tradition (i.e., why not potassium ethoxide and sodium tert-butoxide?) or another reflection of safety issues, it should be noted that alkoxides of different metals tend to have different aggregate structures,[‌16b‌,‌17b‌] that mixed alkoxides involving lithium and heavier alkali metals are known,[‌18b‌] and that, in several systems, different behavior of lithium, sodium, and potassium alkoxides is observed. Thus, the catalytic effects of alkali metal tert-butoxide clusters on the rates of ester interchange reactions are reflected in differences in reactivity as a function of the cation (Li+ < Na+ < K+ < Rb+ < Cs+).[‌19b‌] Furthermore, a new variation on the classical Williamson ether synthesis, involving the preparation of tert-alkyl ethers via the reaction of alkyl halides with alkali metal phenoxides, has shown a variable substitution to elimination ratio depending on the metal.[‌20b‌] Also notable are the differences in the outcome of stereoselective hydrogenation of ketones catalyzed by ruthenium complexes in the presence or absence of tert-butyl alkoxides (sodium or potassium).[‌21b‌,‌22b‌] In another illustration of metal alkoxide reactivity differences, the mechanistically interesting 1,4-dithiin to 1,3-dithiole rearrangement, which proceeds upon deprotonation with alkoxide bases, shows major differences in efficiency depending on the base: potassium tert-butoxide leads to quantitative rearrangement, whereas sodium and lithium tert-butoxides are effective only in the presence of crown ethers and conversions are low.[‌23b‌] A classic and well-recognized reaction involving different reactivity of alkoxides is the anionic oxy-Cope rearrangement where the rates depend on the degree of cation coordination and follow the order K+ > Na+ > Li+.[‌13b‌] The alkyllithium–potassium tert-butoxide combination (Lochmann–Schlosser “superbase”[‌24b‌,‌25b‌]) is highly useful, especially in benzylic and allylic deprotonation reactions (vide infra).[‌26b‌]

Outside the area of synthetic organic chemistry, several important differences in polymerization processes have been identified as resulting from using either sodium or potassium tert-butoxide, for example, one involving a marked change in the tacticity of the polymer.[‌27b‌]

The advent of phase-transfer catalysis (PTC, see Section ‌8.2.4‌) signaled major changes in the development of the chemistry of C—H acids, and progress in the area of enantioselective synthesis have been especially striking.[‌28b‌] While, for good reasons both in terms of concentration and nature of base, 50% sodium hydroxide solution is by far the most commonly used reagent, other systems for phase-transfer catalysis have been devised and developments in the area of solid–liquid phase-transfer catalysis, with applications of solid potassium hydroxide in dimethyl sulfoxide being especially useful, augur well for the future application of phase-transfer catalysis in organic synthesis.[‌29b‌] Although the method now seems to be relegated to textbooks as “classical”, applications abound, e.g. dihalocarbene generation and Wittig reaction (‌Scheme 7‌).[‌30b‌–‌32b‌] Newer extensions include solid supports preloaded with hydroxide, in which potassium hydroxide seems to be emerging as the most promising base.[‌33b‌] In the more classical liquid–liquid media, the replacement of sodium hydroxide by potassium hydroxide can lead to major differences in reactivity and selectivity.[‌34b‌]

‌Scheme 7‌ Use of Phase-Transfer Catalysis in a Wittig Reaction and in Carbene Addition[‌30b‌–‌32b‌]

Following up on the theme of derivatives of alkali metals used as bases, it is noteworthy that the use of unsubstituted sodium and potassium amides as reagents in organic synthesis has nearly disappeared from the literature since the mid-1970s, these strong bases being superseded by lithium dialkylamides, alkyllithium compounds, and, to some extent, sodium and potassium hexamethyldisilazanides. It may be questioned if this is actually scientifically valid and if there is truly no potential in developing dialkylamides of sodium and potassium for organic synthesis. Sodium and potassium derivatives of N—H acids include the amides, hexamethylsilazanides, mono- and disodium cyanamide, and the azides. The amides acting as nucleophiles are encountered in reactions involving alkyl- and arylamines and alkanamides, for example in N-alkylation of such compounds. In this respect, it should be noted that although most of these reactions are simple and can be performed efficiently under phase-transfer catalysis conditions, the choice of base and the experimental protocol may be critical. For example, alkylation of a derivative of tetraazacyclododecane (four N—H centers) with tert-butyl bromoacetate in the presence of potassium carbonate yields the corresponding tetraalkylated derivative, whereas the analogous reaction performed in the presence of solid sodium hydrogen carbonate provides the trialkylated product. These observations were exploited in synthesis of DOTA ligands.[‌35b‌] Significant applications of potassium dialkylamides include the “acetylenic walk” reaction,[‌36b‌] the palladium-catalyzed α-enolate arylation,[‌37b‌] and the sulfur ylide into epoxide transformations[‌38b‌] (‌Scheme 8‌).

‌Scheme 8‌ Alkyne Isomerization and Palladium-Catalyzed Arylation with Potassium Hexamethyldisilazanide[‌36b‌,‌37b‌]

Several product subclass sections give unique overviews of the use of inorganic salts of alkali metals in organic synthesis (Sections ‌8.2.3‌–8.2.6, 8.3.3, 8.4.2, and 8.4.3). Halides of sodium, potassium, rubidium, and cesium, and also cyanides, carbonates, and other salts, appear in a great many diverse reactions and this rich chemistry is difficut to summarize in a concise fashion. For example, sodium halides (Section ‌8.2.3‌8.2.3) are used as nucleophiles in many aliphatic and aromatic substitution reactions, such as the classical Finkelstein reaction (‌Scheme 9‌),[‌39b‌] or the extremely useful dealkoxycarbonylation with sodium chloride (the Krapcho reaction), conditions of which tolerate the presence of many functional groups (‌Scheme 10‌),[‌40b‌] but sodium chloride is also used as the source of molecular chlorine for addition reactions, for reductive dehalogenation of α-halo ketones, and for oxidative transformation of alcohols into carbonyl compounds.

‌Scheme 9‌ The Finkelstein Reaction[‌39b‌]

‌Scheme 10‌ The Krapcho Dealkoxycarbonylation[‌40b‌]

Sodium iodide and sodium bromide are used in combination with transition metals, or their salts, and with lanthanides, or their salts, in a variety of applications, e.g. Suzuki–Miyaura cross coupling[‌41b‌] (‌Scheme 11‌), Heck coupling,[‌42b‌] and indium-mediated allylation of 1,2-diones.[‌43b‌] Selective bromo for iodo exchange in heterocycles using sodium bromide (‌Scheme 12‌),[‌44b‌] and cyano for halo exchange with sodium cyanide[‌45b‌] are widely applicable synthetic processes. Sodium fluoride is well-known as, inter alia, a useful reagent in cleavage of C—Si and O—Si bonds. The classical SN2 displacement of alkyl halides and other corresponding leaving groups is augmented by the possibility to effect the equivalent of this reaction on aryl and vinyl halides under transition-metal catalysis.[‌46b‌]

‌Scheme 11‌ Suzuki–Miyaura Coupling[‌41b‌]

‌Scheme 12‌ Selective Halide Replacement with Sodium Iodide[‌44b‌]

Cesium halides and cesium carbonate are important reagents in organic synthesis with diverse applications, highlights of which include the “cesium templating effect” in macrocyclization, use as activators in numerous reactions, such as addition of nucleophiles to carbonyl groups, Michael and aldol reactions, and transition-metal-catalyzed cross coupling and related reactions, and use of cesium carbonate to generate nucleophiles (Sections ‌8.4.2‌ and ‌8.4.3‌).

In ‌Scheme 13‌, some additional examples of reactions involving alkali metal salts are provided only to pique the reader's curiosity and entice him or her to visit the appropriate sections, since providing a general picture in one scheme is not feasible.

‌Scheme 13‌ Examples of Use of Sodium Halides in Synthesis[‌47b‌–‌52b‌]

Although enjoying relatively special synthetic application (and odors), potassium–sulfur, –selenium, and, much less, –tellurium chemistry finds utility in transformations leading to organochalcogen compounds. Potassium sulfide, selenide, and telluride, as well as the corresponding thiocyanate, selenocyanate, and tellurocyanate, are commercially available or readily prepared reagents. The most widely used reactions of potassium thiocyanate and selenocyanate are alkylations with alkyl halides in which their ambident nucleophilicity to form either C—N or carbon—chalcogen bonds is evidenced,[‌53b‌] in the synthesis of significant heterocycles (‌Scheme 14‌),[‌54b‌] and in stereoselective alkene inversions[‌55b‌] and epoxide to alkene transformations.[‌56b‌] In general, the corresponding sodium–sulfur, –selenium, and –tellurium derivatives are also readily prepared and are, perhaps, more widely used.

‌Scheme 14‌ Heterocycles from Thiocyanate[‌54b‌]

Polar organometallic organosodium and organopotassium compounds derived from sp2 (aromatic, vinyl) and sp3 organic compounds have not, in spite of lower cost, enjoyed general applicability owing to the difficulty in handling of these unstable and pyrophoric substances. The lower stability of sodium and potassium organometallics in solution compared to similar organolithium compounds has been documented.[‌57b‌]

When use of these compounds is considered during synthetic planning “metal effects”, elegantly summarized by Schlosser,[‌57b‌] and also use of organomagnesium compounds (Grignard reagents) must be considered. Briefly, different metals may cause a different reaction to be favored, e.g. proton abstraction over addition to a carbon—heteroatom multiple bond or vice versa, and may be responsible for differences in regio- or stereoselectivity. Mechanistically, the moniker “carbanions” tends to be misleading and a number of mechanisms may operate.[‌57b‌]

While benzyllithiums have seen numerous uses in organic synthesis (cf., Section ‌8.1.13‌), the analogous sodium or potassium species are not popular, perhaps also for instability reasons, and because, under conditions for their generation, side reactions such as ring metalation occur, which can also be a function of the choice of sodium or potassium base used for deprotonation (Section ‌8.2.11‌). Perhaps predictably, benzyllithium, benzylsodium, and benzylpotassium show different reactivity as reflected in yields and selectivities. To illustrate, varying efficiency and diastereoselectivity in reactions of isoindolinones metalated at the benzylic position using either lithium hexamethyldisilazanide, sodium hexamethyldisilazanide, or potassium hexamethyldisilazanide in the synthesis of 3-alkylisoindolin-1-ones has been observed.[‌58b‌]

The important field of directed ortho-metalation of aromatics and heteroaromatics (DoM and HetDoM) has no counterpart in sodium and potassium chemistries due to such species being pyrophoric, unstable, and often unselective, although attempts to change this situation are ongoing.[‌59b‌]

Derivatives of the heaviest alkali metals, cesium and rubidium, show marked differences in chemical behavior and in structure compared to the corresponding lithium compounds (Section ‌8.4.1‌). In ether solvents, fluorenyllithium exists predominantly in the form of solvent-separated ion pairs, whereas analogous cesium compounds show contact ion pair structures. A brief discussion of the aggregation phenomena comparing lithium species with heavier alkali metals derivatives is clearly presented in Section ‌8.4.1‌. Many useful organorubidium and -cesium compounds can be prepared by proton removal from the hydrocarbon by the metal, in contrast to organolithium chemistry where this is usually not promising and the use of alkyllithium or lithium amide bases is required. Synthetic applications of organorubidium and -cesium compounds, such as in addition reactions of these nucleophilic species to carbonyl groups, are not popular (Section ‌8.4.4‌), perhaps due to the much greater instability of such reagents.

In small-scale chemistry, the highly powerful Lochmann–Schlosser superbase combinations are reasonably widely used and a number of synthetic applications have been reported (Section ‌8.3.7‌). Some illustrative examples are shown in ‌Scheme 15‌.[‌60b‌,‌61b‌]

‌Scheme 15‌ Synthetic Applications of Superbases[‌60b‌,‌61b‌]

Organosodium compounds stabilized by an adjacent heteroatom, such as metalated sulfoxides (of which dimsylsodium is the most popular), sulfones, sulfoximines, and nitronates, as well the corresponding α,α-disubstituted species (metalated disulfones, α-nitro sulfones, and dinitronates) have found broad application in organic synthesis (Sections ‌8.2.12‌ and ‌8.2.13‌). The derived species are used in synthetic methods of alkylation or addition to a carbonyl group, or addition–elimination protocols; amongst the latter, the aromatic vicarious nucleophilic substitution (VNS) of hydrogen deserves special mention as it provides a new approach to functionalization of aromatic compounds (‌Scheme 16‌).[‌62b‌] Some reagents in this class provide a convenient source of carbonyl synthetic equivalents with reactivity umpolung; for example methyl (methylsulfanyl)methyl sulfoxide metalated twice (in two separate steps) constitutes a carbonyl dianion equivalent (‌Scheme 17‌).[‌63b‌]

‌Scheme 16‌ The Aromatic Vicarious Nucleophilic Substitution of Hydrogen[‌62b‌]

‌Scheme 17‌ Synthesis of a Ketone by Double Alkylation of Methyl (Methylsulfanyl)methyl Sulfoxide[‌63b‌]

Alkali metal enolates are now well established as the most important class of reactive intermediates for C—C bond formation. Briefly, their reactions include nucleophilic displacement (alkylation, silylation, and replacement of the α-hydrogen with a heteroatom-based functional group, such as hydroxy), addition to carbonyl groups (aldol and related reactions), addition to C=N bonds (Mannich reaction), and addition to electron-deficient C=C bonds (Michael reaction). An abbreviated “cartoon” summary of reaction types is presented in ‌Scheme 18‌. Most of these reaction types have intramolecular variants and many strategies for enantioselective and diastereoselective synthesis of chiral compounds have been developed (see Section ‌8.1.17‌).

In the area of enolate chemistry the sentiment that sodium enolates are superfluous or, at least, overshadowed by lithium enolates surfaces occasionally. The synthetic community would be, however, poorly served if sodium enolates were to be forgotten. The major difference in overall lithium and sodium (or potassium) enolate chemistry is that the former is dominated by kinetically controlled protocols while the latter involves mostly reactions under thermodynamic control.[‌64b‌] The predominance of thermodynamic protocols in chemistry of sodium and potassium enolates is more a consequence of tradition than necessity. Teachers of organic synthesis often observe students falling in love with lithium diisopropylamide and hurriedly suggesting its use for all deprotonation reactions, with the obvious unnecessary complications in such reactions as the venerable malonate and acetoacetate condensations. Bases that are typically used for generation of sodium and potassium enolates involve the corresponding alkali metal hydrides, alkoxides, and hexamethyldisilazanides, the latter allowing the reactions to be performed under kinetic control. Because the hydrides are non-nucleophilic, carrying out certain types of multiple reactions, such as polyalkylation of ketones, in “one pot” by using excess of the base and excess of the electrophile is possible, in contrast to protocols involving lithium amides as bases.[‌65b‌]

There are numerous significant differences in reaction outcomes between lithium, sodium, and potassium enolates including control of O- vs C-alkylation of enolates (sodium enolates give a greater proportion of O- vs C-alkylation; Section ‌8.2.15‌). Another instructive reaction is the alkylation of aldehydes; α-alkylation of aldehydes via deprotonation using lithium amides is not a viable synthetic protocol due to intervening side reactions such as hydride transfer from the amide to the carbonyl group, addition of the amide to the carbonyl group, and, to a smaller extent, self-aldolization,[‌66b‌] but potassium enolates of aldehydes, generated using potassium hydride, can be smoothly alkylated.[‌67b‌] In addition, alkylation and aldol reactions of ketone enolates can differ substantially in regioselectivity depending on the cation (lithium, sodium, or potassium).[‌68b‌,‌69b‌] The enolization process itself can be affected by the choice of base and thus, for example, regioselectivity differences using lithium bases vs sodium hydride or potassium alkoxide bases are known.[‌57b‌] A synthetically very valuable hydroxylation of enolates using 2-sulfonyloxaziridines proceeds in much higher yields with potassium enolates than with lithium enolates.[‌70b‌]

‌Scheme 18‌ Typical Reactions of Alkali Metal Enolates

The importance of structural studies on lithium enolates, especially concerning aggregation phenomena, has led to similar investigation of sodium enolates but not of potassium enolate aggregates. In the solid-state crystal structures, lithium, sodium, and potassium enolates show certain similarities but also differences in aggregation numbers.[‌71b‌,‌72b‌]

This brief overview only touches on some of the multitude of methods and reactions that comprise Volume 8b. To conclude, perhaps an unnecessary statement is that Volume 8, as all others in the Science of Synthesis series, will make its profound impact via the periodic electronic updates. The reader may be richly rewarded in making comparisons especially between the organometallic chemistry of lithium and that of potassium and sodium in solving his or her synthetic problems.

References

References