Aside from the acid chloride substrate, three reagents are required: diazomethane, water, and a metal catalyst. Each has been well investigated. The diazomethane is required in excess so as to react with the HCl formed previously. Not taking
diazomethane in excess results in HCl reacting with the diazoketone to form chloromethyl ketone and N2. Mild conditions allow this reaction to take place while not affecting complex or reducible groups in the reactant-acid. The reaction requires the presence of a
nucleophile (
water). A metal
catalyst is required. Usually
Ag2O is chosen but other metals and even light effect the reaction.
Diazomethane is the traditional reagent, but analogues can also be applied. Diazomethane is toxic and potentially violently explosive, which has led to safer alternative procedures, For example,
diazo(trimethylsilyl)methane has been demonstrated.
Acid anhydrides can be used in place of acid chloride. The reaction yields a 1:1 mixture of the homologated acid and the corresponding
methyl ester. This method can also be used with primary diazoalkanes, to produce secondary α-diazo ketones. However, there are many limitations. Primary diazoalkanes undergo
azo coupling to form azines; thus the reaction conditions must be altered such that acid chloride is added to a solution of diazoalkane and triethylamine at low temperature. In addition, primary diazoalkanes are very reactive, incompatible with acidic functionalities, and will react with activated alkenes including
α,β-unsaturated carbonyl compounds to give
1,3-dipolar cycloaddition products. An alternative to the Arndt–Eistert reaction is the
Kowalski ester homologation, which also involves the generation of a carbene equivalent but avoids diazomethane. ==Reaction mechanism==