Mechanistic Understanding of the Aryl-Dependent Ring Formations in Rh(III)-Catalyzed C–H Activation/Cycloaddition of Benzamides and Methylenecyclopropanes by DFT Calculations

The divergence between Rh­(III)-catalyzed
C–H activation/cycloaddition of phenyl- and 2-furanyl-containing
benzamides with methylenecyclopropanes (MCP) was studied by DFT calculations.
Calculations found that the C–H activation via a CMD mechanism
is the most difficult step of the reaction involving phenyl. In contrast,
the C–H activation of the 2-furanyl-containing substrate is
kinetically easier but the formed five-membered rhodacycle is relatively
unstable, making the following MCP insertion more difficult. Thus,
different KIE data was obtained in experiments. The MCP insertion
forms a seven-membered-ring rhodacycle intermediate, from which the
chemoselectivity of the whole reaction is determined by the competitive
pivalate migration (path I) and β-C elimination (path II). While
the β-C elimination is lower in energy when a furanylene is
contained in the intermediate, a reversed preference of pivalate migration
was predicted for the phenylene counterpart. Structural analysis suggested
that the unfavorable β-C elimination in the phenylene case should
be attributed to the obviously increased ring strain in the corresponding
transition state, instead of the difference in electronic properties
between the aryl groups. This accounts for why aryl-dependent chemoselectivity
was observed. In addition, the results indicated that for both paths
I and II the generation of a Rh­(V)–nitrenoid intermediate from
pivalate migration is crucial for the final C–N bond formation.
This explains why no reaction occurred when the N–OPiv moiety
was replaced with an N–OMe group, as no Rh­(V) intermediate
could be formed in this system.