Radical anion rearrangements. aryl cyclopropyl ketyl anions

Abstract

Aryl cyclopropyl ketones have often been employed as diagnostic probes for single electron transfer (SET) in organic chemical reactions. The implicit assumption in such studies is that the formation of rearranged product(s) can be ascribed to the intermediacy of a ketyl anion. Through a detailed examination of the decay of electrolytically generated aryl cyclopropyl] ketyl anions, we have shown that the assumptions made in the use of these substrates as SET probes are not necessarily valid. Using derivative cyclic and linear sweep voltammetry it was discovered that the ketyl anions of alkyl- and unsubstituted aryl cyclopropyl ketones (class I), including phenyl cyclopropyl ketone (28a), 1-benzoyl-2-methylcyclopropane (28b), 1-benzoyl-2,2-dimethylcyclopropane (28c), p-tolyl cyclopropyl ketone (28d), and 1-benzoyl-1-methylcyclopropane (28e), undergo a slow and reversible cyclopropyl carbinyl type rearrangement followed by dimerization of the ring-opened and ring-closed radical anions. The equilibrium constant for the reversible ring opening lies highly in favor of the ring closed form. For (28a^·⁻) in anhydrous N,N-dimethylformamide containing 0.5 M n-Bu₄NBF₄ at 23 °C, the equilibrium constant was estimated at K ≈ 4.6 x 10⁻⁸ with a maximum rate constant for ring opening and a minimum rate constant for ring closing at 2.0 s⁻¹ and 4.3 x 10⁷ s⁻¹ respectively; the rate constant for dimerization was placed at 8.4 x 10⁷ M⁻¹s⁻¹. Semiempirical molecular orbital calculations (AM1) complement the above observations.

Similar results were obtained for all class I compounds. The ketyl anions of aryl cyclopropyl ketones with good radical stabilizing groups on the cyclopropane ring (class ID), including trans-1-benzoyl-2-phenylcyclopropane (66), and 1-benzoyl-2-vinylcyclopropane (76) undergo rapid unimolecular ring opening. The rate constants for opening of (66^·⁻) and (76^·⁻) are greater than 10³ s⁻¹ but probably less than 10⁷ s⁻¹ and 10⁵ s⁻¹ respectively. Based upon our findings, class I ketones are extremely unreliable probes for SET; class II ketones may prove to be useful SET probes, but since absolute rate constants for their rearrangement are not yet known, they should be used only with extreme caution. The implications of these results are discussed in light of utilizing aryl cyclopropyl ketones as probes for SET.