Introduction

Jonathan Clayden    Department of Chemistry, University of Manchester, Manchester, UK

1.1 Scope and Overview

Organolithiums are central to so many aspects of synthetic organic chemistry that a book on organolithium chemistry must be a book on synthesis. Hardly a molecule is made without a bottle of BuLi: evaporation as butane is the destiny of at least one proton of the starting material in almost any synthetic sequence.

A book on organolithiums is also a book on mechanism. The observation of selectivity in an organolithium reaction has often led to mechanistic insights that later turn out to be general mechanistic features of organic reactions. Directed metallation, for example, started with the ortholithiation of anisole, and led to directed reactions of zinc and palladium. A decade of close investigation of configurational stability in the organolithium series preceded similar studies on organozincs and other organometallics. “Dynamic thermodynamic resolution” is a mechanistic feature first identified in the reactions of organolithiums in the presence of sparteine, and later exploited right outside of the organometallic sphere, in the synthesis of atropisomers.

Given their ubiquity, this book concentrates on one feature of the reactions of organolithiums (pushed to a wider sense in some areas thari others) selectivity. This feature is the result of the civilisation of organolithiums from the savage beasts of 40 years ago (BuLi, benzene, reflux) to the tamed, well-trained species we use to coax out one proton at a time or to nudge a starting material over the energetic barrier of a spectacular cascade reaction.

To explain selectivity I have discussed mechanism in detail but not depth – the mechanistic discussions are intended not as a full account of current understanding of organolithium structure and reactivity, but as a tool for use in predicting likely outcomes of reactions and in accounting for unlikely ones. Similarly, structure is dealt with where necessary to explain a point, but detailed discussions of organolithium structure is outside the scope of a book primarily about organolithium reactions. I have also limited the definition of “organolithium” to those compounds in which there is a clear C–Li bond: compounds with any degree of enolate structure, and lithiated sulfones, sulfoxides, phosphonates, phosphine oxides etc. have been excluded. Inclusion or exclusion of a compound should not be taken to imply anything about its structure – a limit had to be drawn somewhere, and in some discussions the limit is stretched further tan in others.

General points about organolithiums in solution are considered briefly first, followed by chapters addressing the synthesis of functionalised organolithiums, and in particular the various methods for achieving regioselectivity. Some organolithiums have stereochemistry, and an account of the stereoselective synthesis of organolithiums follows an explanation of which and why. Discussions of reactions follow, but rather than give a thin account of a wide range of well-known additions and substitutions, I have limited the coverage to some important and developing areas: stereospecific and selective reactions, particularly those involving (–)-sparteine and other chiral additives, inter- and intramolecular additions to π systems, and rearrangements.

1.2 Organolithiums in solution

Organolithiums (with the exceptions of methyllithium and phenyllithium) are remarkably soluble even in hydrocarbon solvents,1,2 and simple organolithium starting materials are available as stable hydrocarbon solutions (Table 1.2.1). Methyllithium and phenyllithium are indefinitely stable at ambient temperatures in the presence of ethers, and are solubilised by the addition of ether or THF.

Hydrocarbon solutions of n-, s- and t-BuLi are the ultimate source of most organolithiums, but a number of other bases are widely used to generate organolithiums from more acidic substrates. Among these are LDA, LiTMP and other more hindered lithium amide bases, and hindered aryllithiums such as mesityllithium and triisopropylphenyllithium.

The electron-deficient lithium atom of an organolithium compound requires greater stabilisation than can be provided by a single carbanionic ligand, and freezing-point measurements indicate that in hydrocarbon solution organolithiums are invariably aggregated as hexamers, tetramers or dimers.3 The structure of these aggregates in solution can be deduced to a certain extent from the organolithiums’ crystal structures4 or by calculation:5 the tetramers approximate to tetrahedra of lithium atoms bridged by the organic ligands; the hexamers approximate to octahedra of lithium atoms unsymmetrically bridged by the organic ligands.6,7 The aggregation state of simple, unfunctionalised organolithiums depends primarily on steric hindrance. Primary organolithiums are hexamers in hydrocarbons, except when they are branched β to the lithium atom, when they are tetramers. Secondary and tertiary organolithiums are tetramers, while benzyllithium and very bulky alkyllithiums (menthyllithium) are dimers.1,3

Coordinating ligands – such as ethers or amines, or even metal alkoxides (see section 2.6) – can provide an alternative source of electron density for the electron-deficient lithium atoms. These ligands can first of all stabilise the aggregates by coordinating to the lithium atoms at their vertices, and then allow the organolithiums to shift to an entropically favoured lower degree of aggregation. As shown in Table 1.2.3, the presence of ether or THF typically causes a shift down in aggregation state, but only occasionally results in complete deaggregation to the monomer.1 Methyllithium, ethyllithium and butyllithium remain tetramers in Et2O, THF,8 or dimethoxyethane (DME),9 with some dimer forming at low temperatures;10 t-BuLi becomes dimeric in Et2O and monomeric in THF at low temperatures.11 In the presence of TMEDA alone, however, s-BuLi remains a tetramer.12