Substrate-adaptive reaction development enables modular access to functionalized activated olefins
Year: 2026
Type: preprint
Abstract: The decarboxylative olefination of aldehydes with malonic acid derivatives offers a mild and modular route to activated olefins, allowing the β-substituent, α-substituent and carbonyl-derived activating group to be varied independently. However, this modularity has not been realized within a unified framework: Existing methods remain fragmented across individual carbonyl classes and are dominated by αunsubstituted products, while access to α-heteroatom-substituted olefins, particularly α-halo derivatives, remains poorly developed. Moreover, these methods provide little guidance for adapting the reaction to new or challenging substrates. Consequently, despite its attractive features, decarboxylative olefination remains underutilized relative to established methods such as the Horner–Wadsworth–Emmons reaction. Here, we develop a unified, substrate-adaptive platform for the synthesis of functionalized activated olefins. Rather than optimizing a single model reaction and applying the resulting conditions across a fixed substrate scope, we used Bayesian optimization across representative substrate pairs to generate a compact, information-rich reaction-development dataset. This dataset was then used directly as prior knowledge for active transfer learning, enabling the identification of tailored, high-yielding conditions for new substrate combinations and complex targets with minimal additional experimentation. The resulting workflow provides access to α,β-unsaturated thioesters, esters and amides, substantially expands the scope of α-functionalized olefins and identifies effective conditions for substrates that were limitations of previous fixed-condition substrate scopes. More broadly, this study demonstrates that reactiondevelopment datasets can serve as transferable reaction knowledge, enabling efficient substrate-adaptive optimization and accelerating the maturation of emerging synthetic methods into practically general transformations.
Source: ChemRxiv
Institution: New York University Abu Dhabi
Language: English
Subfield: Organic Chemistry
Field: Chemistry
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7204955512