Generalizing arene C–H alkylations by radical–radical cross-coupling

The efficient and modular diversification of molecular scaffolds, particularly for the synthesis of diverse molecular libraries, remains a notable challenge in drug optimization campaigns^{1, 2–3}. The late-stage introduction of alkyl fragments is especially desirable due to the high sp³ character and structural versatility of these motifs⁴. Given their prevalence in molecular frameworks, C(sp²)–H bonds serve as attractive targets for diversification, although this process often requires difficult prefunctionalization or lengthy de novo syntheses. Traditionally, direct alkylations of arenes are achieved by using Friedel–Crafts reaction conditions with strong Brønsted or Lewis acids^5,6. However, these methods suffer from poor functional group tolerance and low selectivity, limiting their broad implementation in late-stage functionalization and drug optimization campaigns. Here we report the application of a new strategy for the selective coupling of differently hybridized radical species, which we term ‘dynamic orbital selection’. This mechanistic model overcomes common limitations of Friedel–Crafts alkylations via the in situ formation of two distinct radical species, which are subsequently differentiated by a copper-based catalyst on the basis of their respective binding properties. As a result, we demonstrate here a general and highly modular reaction for the direct alkylation of native arene C–H bonds using abundant and benign alcohols and carboxylic acids as the alkylating agents. Ultimately, this solution overcomes the synthetic challenges associated with the introduction of complex alkyl groups into highly sophisticated drug scaffolds in a late-stage fashion, thereby granting access to vast new chemical space. Based on the generality of the underlying coupling mechanism, ‘dynamic orbital selection’ is expected to be a broadly applicable coupling platform for further challenging transformations involving two distinct radical species.