Mechanism of Nucleophilic Addition to the Carbonyl Group

Every reaction discussed in the rest of this chapter's aldehyde/ketone sections -- addition of $\text{HCN}$, of bisulphite, of alcohols, of Grignard reagents, of amines -- follows the same general mechanistic pattern, because all of them exploit the same structural feature described in §8.2: an electron-deficient, planar, $sp^2$ carbonyl carbon.

Step 1: nucleophilic attack. A nucleophile $\text{Nu}^-$ (or a neutral nucleophile with a lone pair, such as an amine) approaches the carbonyl carbon along a trajectory roughly perpendicular to the plane of the $\pi$ bond, rather than straight along the $\text{C=O}$ axis -- this angle of approach allows the nucleophile's lone pair to overlap efficiently with the empty, low-energy $\pi^*$ antibonding orbital of the carbonyl group as the new bond forms. As the new $\text{C-Nu}$ $\sigma$ bond begins to form, the $\pi$ electrons of the $\text{C=O}$ bond are pushed entirely onto the oxygen, converting the carbonyl oxygen from a neutral, doubly-bonded atom into a singly-bonded, negatively charged alkoxide oxygen.

Step 2: the tetrahedral intermediate. The carbonyl carbon, which was $sp^2$ and trigonal planar before attack, rehybridises to $sp^3$ as the new bond forms, becoming a tetrahedral centre bonded to four groups: the incoming nucleophile, the (now negatively charged) oxygen, and the two original substituents that were already on the carbonyl carbon. This tetrahedral alkoxide intermediate is the common branch point of the whole family of addition reactions -- what happens to it next (simple protonation, loss of water to reform a different double bond, or displacement of a leaving group) is what distinguishes one named reaction from another later in the chapter.

Step 3: protonation (or an equivalent step). In the simplest case, the alkoxide oxygen of the tetrahedral intermediate is protonated -- by the solvent, by trace acid, or on aqueous work-up -- to give the neutral tetrahedral addition product, with the original $\text{C=O}$ now replaced by a single $\text{C-OH}$ (or, when the nucleophile was an amine, by $\text{C-NH-R}$, which can go on to lose water in a further step to give an imine; see §8.6). …