Unlocking Direct C–S Bonds: Radical C–SuFEx Ligation of Sulfonyl and Sulfonimidoyl Fluorides with Alkenes and Alkynes

Radical C–SuFEx Ligation of Sulfonyl and Sulfonimidoyl Fluorides with Alkenes and Alkynes: Our team recently collaborated with the research groups of Prof. Gui Lu and Prof. Jiang Weng from the School of Pharmaceutical Sciences at Sun Yat-sen University to develop a novel sulfur-fluorine exchange (SuFEx) reaction strategy via visible-light photoredox catalysis. This work has been published in Angewandte Chemie International Edition.

This methodology achieves the construction of carbon-sulfur (C–S) bonds from hexavalent sulfur-fluorine (S(VI)–F) compounds under mild conditions for the first time, overcoming a long-standing scientific bottleneck: the inability of traditional SuFEx reactions to directly form S(VI)–C bonds. This discovery not only expands chemical space—offering a new approach for discovering and optimizing sulfur-containing drug molecules—but also significantly enhances our core offerings: the Innovative Building Block Library and the VAST virtual compound library.

Efficient C–SuFEx Ligation Achieved Under Mild Conditions

As a next-generation “click chemistry” tool, sulfur-fluorine exchange (SuFEx) chemistry has shown tremendous potential in materials science and bioorthogonal fields. However, current SuFEx methodologies primarily employ O- and N-nucleophiles to afford S–O and S–N bonds, while reports on C-nucleophiles for S–C bond formation have remained limited.

To address this, our team, in collaboration with  Sun Yat-sen University, innovatively introduced a visible-light photoredox catalysis strategy. By forming an electron donor-acceptor (EDA) complex between triarylamines and hexavalent sulfur-fluorine compounds, the team efficiently activated the S(VI)–F bond under neutral, room-temperature conditions. This process generates key radical intermediates to successfully achieve the radical C–SuFEx reaction. The method allows for the highly efficient synthesis of a series of sulfur-containing functional molecules, such as alkynyl sulfones, alkenyl sulfones, and alkenyl sulfonimide derivatives, and demonstrates excellent functional group tolerance and substrate scope (Scheme 1).

Scheme 1. The scope of the radical SuFEx reactions of sulfonyl fluorides with alkynes and alkenes

Notably, this methodology was successfully applied to the late-stage modification of complex natural products, approved drugs (including Celecoxib, Valdecoxib, and Isoxepac), and steroid-containing molecules, highlighting its immense value in drug discovery. Using quantum mechanics (QM) calculations, we precisely predicted substrate reactivity and S–F bond cleavage modes, providing theoretical guidance for the experimental design. This closed-loop model of “computation guiding experiments, and experiments validating computation” represents XtalPi’s core competitiveness. Future methodological R&D will increasingly leverage AI Agents and autonomous experimentation to achieve even greater automation.

The discovery directly empowers our platforms to accelerate sulfur-containing drug discovery:

1. Innovative Building Block Library: Enriching Sulfur-Containing Chemical Space

Due to their unique physicochemical properties and biological activities, sulfur-containing building blocks are in high demand in drug molecular design. However, traditional synthetic methods suffer from limited production capacity. This technology enables complex sulfur-containing building blocks that were previously difficult to synthesize (such as structures containing alkynyl sulfones and alkenyl sulfones) to be efficiently constructed under mild conditions. We will soon integrate this method into our building block library to provide global clients with highly differentiated, high-value sulfur-containing building blocks, accelerating the generation and optimization of lead compounds.

2. VAST Agent

Based on our proprietary AI models, we have built the VAST compound library. Different subsets of the VAST library contain billions to trillions of synthesizable molecules. The accessibility of each molecule has been precisely predicted via AI models, achieving a synthesis success rate of over 85% and representing a 30% increase in synthesis efficiency. To date, dozens of pharmaceutical companies have utilized VAST for hit discovery. This new methodology will expand the chemical space of VAST, further accelerating first-in-class drug discovery.