Catalytic Cycle for the First Stage

The catalytic cycle for the formation of 3PN and 2M3BN is shown in Fig. 7.13. The following points deserve attention. First, interactions of Lewis acid with the coordinated cyano groups are not shown. The evidence for such interactions comes from full characterization of HNiL3CN-BPh3 (L = o-tolylphosphite) as well as detailed IR and multinuclear NMR studies of a number of analogous nickel complexes. Second, the intermediates NiL3 and 7.45- 

47 are well characterized by IR and multinuclear NMR techniques. Single crystal X-ray structures of NiL2 (alkene), where the alkene is ethylene or acrylonitrile and L is o-tolyl phosphite, have been determined. These are obvious models for the proposed intermediates 7.48 and 7.50. As both these intermediates are 16-electron species, addition of L to give 7.49 or 7.51 prior to the elimination of 3PN or 2M3BN are reasonable mechanistic steps. 

Although for clarity the reactions of Fig. 7.13 are shown to be unidirectional, all the reactions of the catalytic cycle are in fact reversible. This is an important aspect of the first stage of the hydrocyanation process. It provides for a mechanism for the isomerization of the unwanted 2M3BN to the desired 3PN. The isomerization reaction of 2M3BN to 3PN has been studied by deuteriumlabeling experiments. The results are consistent with a mechanism where butadiene is formed in one of the intermediate steps. This means that the reversibility of all the steps allows isomerization to follow the path 7.51 — 7.50 —  

Note that dehydrocyanation of 2M3BN (i.e., 7.50 to 7.47) is nothing but a simple oxidative addition reaction. This is shown formally by reaction 7.21. Reaction 7.22 shows the formal mechanism of butadiene formation and conversion of 7.47 to 7.46. Between the two isomers, 3PN is thermodynamically more stable than 2M3BN. A mixture of these two nitriles, if allowed to reach 

---