| Résumé | Solvent effects on the kinetics of the nitroxide radical trapping (NRT) of various carbon-centered radicals have been probed both by using the radical "clock" method and by the laser flash photolysis (LFP) technique. Although the rate constants for NRT, kT, are lower than the diffusion-controlled limit they are, nevertheless, influenced by solvent viscosity. Rate constants arc even more strongly influenced by the ability of the solvent to solvate the nitroxide. Thus, using the 2,2-dimethyl-to 1,1-dimethyl-3-butenyl radical clock rearrangement, 1* → kC 2*, at 80 °C (kc = 2.4 × 107 s-1) with 1,1,3,3-tetramethyl-isoindole-2-oxyl (TMIO) as the trap, in 32 solvents ranging from to aqueous methanol, it found that log (KT/kT)/M-1) was strongly correlated with the nitroxide's solvation, as gauged by the solvent's effect on the nitrogen hyperfine splitting of a structurally analogous nitroxide ((r) = 0.961 for 26 nonhydroxylic solvents, the hydroxylic solvents forming a separate group). Similar results were obtained at 80 °C with five other radical clocks using smaller solvent sets. Comparison of these radical clock data with the kinetic results obtained by LFP at 18 ° for the NRT of benzyl (22 solvents), n-nonyl (4), and neopentyl (6) by Tempo provides the first unequivocal proof that the kinetics of commonly used alkyl radical clock rearrangements are essentially uninfluenced by solvent properties. Although NRT is primarily an activation-controlled reaction, the magnitude of kT is decreased by an increase in solvent viscosity as is clearly indicated by LFP data for the trapping of benzyl radicals by the sterically unencumbered, Bredt's rule protected nitroxide, 9-azabicyclo[3.3.1]nonane-N-oxyl (ABNO) in saturated hydrocarbons (η = 0.3-16 cP). Using a model for a partially diffusion-controlled reaction, we obtained a theoretical diffusion-controlled limiting rate constant, ks ≈ 3.5 × 109 M-1 s-1, for ABNO/benzyl coupling in a solvent of viscosity η = 1 versus an extrapolated zero viscosity or "activation" limit, k∞ = 1.4 × 109 M-1 s-1. The Tempo/benzyl coupling in saturated hydrocarbons is less curtailed by diffusion since the diffusion/activation ratio is higher, viz., ks/k∞ ≈ 3.0 × 109/0.48 × 1091 (for η = 1). |
|---|
