Unimolecular net heterolysis of symmetric and homopolar σ-bonds

Abstract

The unimolecular heterolysis of covalent σ-bonds is integral to many chemical transformations, including SN1-, E1- and 1,2-migration reactions. To a first approximation, the unequal redistribution of electron density during bond heterolysis is governed by the difference in polarity of the two departing bonding partners1–3. This means that if a σ-bond consists of two identical groups (that is, symmetric σ-bonds), its unimolecular fission from the S0, S1, or T1 states only occurs homolytically after thermal or photochemical activation1–7. To force symmetric σ-bonds into heterolytic manifolds, co-activation by bimolecular noncovalent interactions is necessary4. These tactics are only applicable to σ-bond constituents susceptible to such polarizing effects, and often suffer from inefficient chemoselectivity in polyfunctional molecules. Here we report the net heterolysis of symmetric and homopolar σ-bonds (that is, those with similar electronegativity and equal leaving group ability3) by means of stimulated doublet–doublet electron transfer (SDET). As exemplified by Se–Se and C–Se σ-bonds, symmetric and homopolar bonds initially undergo thermal homolysis, followed by photochemically SDET, eventually leading to net heterolysis. Two key factors make this process feasible and synthetically valuable: (1) photoexcitation probably occurs in only one of the incipient radical pair members, thus leading to coincidental symmetry breaking8 and consequently net heterolysis even of symmetric σ-bonds. (2) If non-identical radicals are formed, each radical may be excited at different wavelengths, thus rendering the net heterolysis highly chemospecific and orthogonal to conventional heterolyses. This feature is demonstrated in a series of atypical SN1 reactions, in which selenides show SDET-induced nucleofugalities3 rivalling those of more electronegative halides or diazoniums.