Reduction with Diimide

Pasto, Daniel J.; Taylor, Richard T.


The reduction of a double bond in the presence of hydrazine appears to have been first observed in 1905 during the reaction of glyceryl oleate, which produced stearic hydrazide. That hydrazine could act as a reagent for the reduction of a carboncarbon double bond was firmly established much later, at which point it was shown that oleic acid could be reduced to stearic acid by treatment with hydrazine, or with hydrazine and sulfur. In 1941 it was reported that vinyl groups in chlorins and porphyrins are selectively reduced to ethyl groups by hydrazine under mild conditions. The synthetic potential of this type of reduction was not recognized until the early 1960s when results from several independent laboratories implicated diimide (HNNH) as the actual reducing agent.

Evidence for the existence of diimide (diimine and diazene) was first obtained in 1892 in the decarboxylation of dipotassium azodicarboxylate, which produced equimolar quantities of nitrogen and hydrazine by the proposed disproportionation of diimide. In 1910 it was proposed that diimide was formed in the reaction of benzenesulfonylhydrazide with hot alkali. Following the proposal that diimide is the reactive intermediate in these reduction reactions, numerous experimental and theoretical studies were launched to find other methods for the synthesis of diimide and to determine the structure(s) of the reactive intermediate(s) and the mechanism of the reduction reaction. An excellent review has appeared which covers the literature on the structure and molecular properties, spectral characterization, and gas-phase reactions of diimide. In this chapter only highlights of such areas are covered. Two reviews covering reductions with diimide appeared in 1965, but none since that time. Most organic texts describe diimide reductions, but not in significant detail.

There are three potential structures for diimide: cis– and trans-diimide and 1,1-diimide (aminonitrene).

trans-Diimide can be generated and trapped at low temperature by a gas-phase electric discharge in hydrazine and by the thermal decomposition of metal salts of p-toluenesulfonylhydrazide. Recently, 1,1-diimide has been generated and trapped by the low temperature photochemical decomposition of carbamoyl azide. Although cis-diimide must be formed as a reactive intermediate in many systems, it has not yet been unambiguously characterized.

The diimide system has been subjected to several theoretical studies at many different basis set levels.

The results of stereochemical studies on the reduction of alkenes and alkynes have led to the suggestion that cis-diimide is the reactive hydrogen-transfer reagent. These results support the suggestion that cis-diimide is the active hydrogen-transfer reagent. The fact that cis-diimide has not been observed and that the calculated inversion and rotation barriers are too large to provide a rate of isomerization of trans– to cis-diimide that would be sufficiently high to account for the observed rate of reduction provides for a mechanistic dilemma. In gas-phase reactions isomerization of trans– to cis-diimide has been proposed to be the rate-limiting step. In solution, however, the isomerization in all probability occurs via a catalyzed process, probably involving a rapid protonation–deprotonation sequence. In this chapter the use of the term “diimide” implies cis-diimide as the reducing agent.

The energy barriers for the disproportionation of cis– with cis-, and cis– with trans-diimide are calculated to be 19.3 and 23.8 kcal per mole, considerably smaller than the barriers for hydrogen transfer to a carboncarbon double bond. From a practical point of view, this competing disproportionation requires the use of considerable excesses of the diimide precursors in the reduction reactions.