The coupling of an organometallic and an organic halide should be a useful way of forming a C-C single bond. The reaction is, however, not general, but restricted to special cases. Grignard reagents will react efficiently with some halides, such as allyl halides, while copper reagents can be used in this type of reaction with a wider range of substrates. Even so, that range is not great and copper reagents are thermally unstable, as well as sensitive to air and moisture.A useful method would not require special conditions, be quite general and use easily prepared starting materials. This can be done with transition-metal catalysis (Scheme 2.1). By far the most widely used metal is palladium, but other metals, especially nickel and, more recently, iron, have also been employed. 1 The main variables to consider are the "R" and "R " groups that can be employed, the main group metals "M" that can be used, the nature of the leaving group "X" and the identity and quantity of the ligand "L". The basic mechanism for palladium-catalysed coupling is a simple combination of oxidative addition, transmetallation and reductive elimination (Scheme 2.2).A coordinatively unsaturated palladium(0) complex 2.1, which is the catalytic species, is generated from an 18-electron palladium(0) complex by reversible ligand dissociation. Alternatively, it can be generated by reduction of a palladium(II) complex. Both the L 4 Pd and the L 2 PdX 2 should be regarded as "pre-catalysts". The catalyst or pre-catalyst complex may be assembled in situ from a mixture of a palladium source, such as a palladium(II) salt, Pd 2 (dba) 3 or even, in a few cases, palladium on carbon, and the ligand. This avoids the need to preform a palladium complex.If a palladium(II) pre-catalyst is employed, then there must be a reduction step prior to the start of the catalytic process. This has sometimes been done by the addition of a reducing agent, but more commonly, the palladium(II) is reduced by homocoupling of the organometallic partner, RM (Scheme 2.3). This homocoupling reaction is of occasional synthetic use itself (see Section 2.10).So how do we know which catalyst to employ? It is determined by looking for precedent and by doing experiments. Key variables are the ligand to metal ratio, and the identity of the ligand. Historically, Ph 3 P has been the most widely used, but that doesn't mean it is best (although it is cheapest). A less widely explored variable is the metal.