The mechanism of CF2-transfer from TMSCF3 (1), mediated by TBAT (2-12 mol%) or by NaI (5-20 mol%), has been investigated by in situ / stopped-flow 19 F NMR spectroscopic analysis of the kinetics of alkene difluorocyclopropanation, and competing TFE / c-C3F6 / homologous perfluoroanion generation, 13 C/ 2 H KIEs, LFERs, CF2-transfer efficiency and selectivity, the effect of inhibitors, and density functional theory (DFT) calculations. The reactions evolve with profoundly different kinetics, undergoing auto-inhibition (TBAT) or stochastic auto-acceleration (NaI), and co-generating perfluoroalkene side products. An overarching mechanism involving direct and indirect fluoride transfer from a CF3-anionoid to TMSCF3 (1) has been elucidated. It allows rationalization of why the NaI-mediated process is more effective for less-reactive alkenes and alkynes, why a large excess of TMSCF3 ( 1) is required in all cases, and why slow-addition protocols can be of benefit. Issues relating to exothermicity, toxicity, and scale-up are also noted.a Relative rates (krel) are for competitive first-order CF2 capture by the alkene/yne, not to overall rates of reaction. b Values in parenthesis by DFT. 26 See sections S3.7, S3.8 and S6.2 in the SI.
Singlet CF2 as the ReactiveIntermediate. We began by studying the reaction of TMSCF3 (1) with alkenes 3i, and E/Z-4 and alkyne 5, Table 1. All underwent difluorocyclopropanation, to varying degrees of conversion, in the presence of TMSCF3 (1, 1.5 M) and 1-5 mol % TBAT, or NaI. Reactions of E/Z-4 proceeded stereospecifically, and with >98 % retention. The difluorocyclopropene 8, generated in low yield (12%) from alkyne 5, under the TBATmediated conditions, underwent partial decomposition to unidentified products. In contrast, 8 was quantitatively-generated, and stable, under NaI-mediated conditions, see section S3.3 in the SI. The same difluorocyclopropane products (6, 7) were obtained from 3i and E/Z-4 on thermalization with the zwitterionic CF2-source Ph3PCF2CO2. 22 The relative reactivities (krel) of alkenes 3i, E-4, and Z-4, and the LFER correlation for a-methylstyrenes (3i-vii, r + = -0.6), 23 were independent of the reagent (1 / Ph3PCF2CO2), and initiator (TBAT / NaI), Table 1, 24 within experimental error. Scheme 3. Experimental a and calculated b KIEs for rapid addition of transient singlet CF2 to 3i, at 300 K. 25 a Experimental (exp.) values in THF; and in PhCl, as solvent. 25 b Calculated (calc.) values by DFT, at the M06/6-31+G* level in Gauss-ian09 employing IEF-PCM single points to account for solvation, and goodvibes, kinisot and PyQuiver to compute free energy corrections and KIEs, see sections S1.6 and S6.2 in the SI. 26Kinetic isotope effects for the reaction of p-F-a-methylstyrene 3i with TMSCF3 (1) initiated by TBAT were obtained by a series of competitions of 13 C-and 2 H-labelled a-methylstyrenes 3i against aryl-D4-3i, monitored by 19 F NMR spectroscopy (aryl-DdF = 0.5 ppm). 25 The resulting primary and secondary kinetic isotope effects, Scheme 3, were consistent with thos...