Computational Mechanism Study of Catalyst-Dependent Competitive 1,2-C→C, −O→C, and −N→C Migrations from β-Methylene-β-silyloxy-β-amido-α-diazoacetate: Insight into the Origins of Chemoselectivity

Abstract

Doyle et al. [J. Am. Chem. Soc. 2013, 135, 1244−1247] recently reported an efficient catalyst-controlled chemoselectivity of competitive 1,2-C→C, −O→C, and −N→C migrations from β-methylene-β-silyloxy-β-amido-α-diazoacetates using dirhodium or copper catalysts. With the aid of density functional theory calculations, the present study systematically probed the mechanism of the aforementioned reactions and the origins of the catalyst-controlled chemoselectivity. Similar to the method reported in the literature, simplified catalyst models Rh2(O2CH)4 and Rh2(N-methylformamide)4 have been used in our initial calculations. However, using the Rh2(O2CH)4 model could not describe the energies of all possible pathways, and high selectivity of three competitive migrations could not be achieved. In order to appropriately describe this 1,2-migration system, real catalyst models Rh2(cap)4, Rh2(esp)2, and CuPF6 have been employed. It was found that the steric and electronic effects of ligands significantly influence the free energy barrier, which ultimately changes the chemoselectivity. In the CuPF6 system, the electronic effects, coupled with the steric factor, give a qualitative explanation for the exclusive chemoselectivity of 1,2-N→C migration over 1,2-C→C or −O→C migration. On the other hand, the bulky ligands of dirhodium catalysts result in the significant steric hindrance around the dirhodium centers and withdrawal of the empty space around the bulky −OTBS group. By analyzing the divergence of three different migration transition states using the distortion/interaction and natural bond orbital analyses, it was found that the 1,2-N→C migration will suffer from a high free energy barrier because of the steric repulsion between the carbonyl group and the carbonyl oxygen of the pyrazolidinone ring. For 1,2-C→C and −O→C migrations, changing the ligands of dirhodium catalysts can change the electronic properties of carbenes, and that is the reason for controlling the major product by changing the dirhodium catalysts. The mechanistic proposal is supported by the calculated chemoselectivities, which are in good agreement with the experimental results.