To further cast light on the mechanism, theoretical calculations were performed at the density functional theory level (B3LYP). For the convenience of calculation, the active catalyst Cp
*Rh(OAc)
2 was chosen as the starting point (zero value of energy). Using
N-methoxy benzamide
1a as a substrate, N-H deprotonation followed by C-H activation were performed
via a concerted metalation-deprotonation (CMD) mechanism with acetate acting as intramolecular base, through transition states
TS-1 (Δ
G‡ = 13.0 kcal/mol) and
TS-2 (Δ
G‡ = 19.8 kcal/mol), respectively (
Fig. 2). The intermediate verification experiments in
Scheme 7c echoed the calculation results. Thereafter, the insertion of
gem-difluorocyclopropene
2ai into the rhodacycle
INT-5 presented two characteristic spatial arrangements,
TS-3 (Δ
G‡ = 27.5 kcal/mol) and
TS-3′ (Δ
G‡ = 28.9 kcal/mol), both of which had a higher activation barrier than the first two steps (
Fig. 3). The computational results indicated that C-H activation was not the turnover limiting step in the reaction, consistent with the observed small experimental KIE values (
Scheme 7b). Taking into account the higher energy barrier of
TS-3′, especially
TS-4′, therefore subsequent calculations revolved around
TS-3. From
INT-7, the priority of either
β-fluorine elimination or C-N bond formation was discussed. The results revealed that the direct
β-fluorine elimination with or without the assistance of acetic acid, followed by C-N bond formation step
via TS-4a and
TS-4b, featured a high energy barrier of 33.2 and 35.1 kcal/mol, reaspectively. Two possible pathways for C-N bond formation prior to the defluorination were then calculated. Considering the high energy barrier of
TS-4c (Δ
G‡ = 68.6 kcal/mol), the direct migration of the methoxy group from the amide to the trivalent rhodium to form
INT-9c was tough. The migration process was more reasonable in the assistance of acetic acid, because the energy barrier of
TS-4 was reduced to 28.8 kcal/mol and a Rh(Ⅴ) intermediate
INT-10 was produced with the free-energy of −54.3 kcal/mol. the synergistic effect of rhodium and acetate accelerated the ring-opening defluorination of
INT-10 to release the final product
3ai (Δ
G‡ = −71.0 kcal/mol). Overall, the computed Gibbs free-energy changes of the reaction pathway demonstrated a redox-neutral Rh(Ⅲ)-Rh(Ⅴ)-Rh(Ⅲ) catalytic cycle for the developed protocol involving HOAc-prompted oxidative addition and unprecedented C-F bond cleavage/ring expansion processes.