Abstract View

Objectives: Dibenzothiophene sulfonium salt (DBTS) recently emerged as a new class of precursors for radiofluorination.^[1] Although this approach has been applied to the production of a handful radiotracers, including for routine clinical production,^[2] the versatility of the method has never been entirely investigated. Here we propose a comprehensive exploration of the DBTS tolerance of various functional groups, solvents, radioprotectants, and water content, adapting a High Throughput (radio)Synthesis (HTS) methodology and [¹⁸F]FPEB as model substrate.^[3]

Methods: All radiofluorinations were carried out on a 24-wells HTS platform. Briefly, a solution of base and cryptand (e.g. K₂₂₂/KHCO₃) in a water/acetonitrile was added in 1 mL glass vials and solvents were removed. To these vials were added acetonitrile solutions of 1) a contaminant, 2) [¹⁸F]FPEB DBTS precursor and 3) azeotropically dried [¹⁸F]fluoride (approx. 5 MBq). Reactions were heated to 80 °C for 5 min before being quenched with 20% EtOH in H₂O. Crude reaction mixtures were subsequently analyzed by radio-TLC to determine radiochemical conversion. (RCC). The ratio between RCCs and the RCC of control experiment (i.e. no contaminant) was calculated to determine the effect (detrimental, neutral, beneficial) of the contaminant.

Figure 1. Overview of the HTS pipeline to explore the landscape of ¹⁸F-radiolabelling with sulfonium salt precursors.

Results: Three different methods were tested to dry the cryptand/base solution for the HTS: under vacuum, under a flow of nitrogen or in an oven. The latter proved to be the fastest, the most efficient and reproducible, with control experiments yielding RCCs of 50±9% (n = 32). DBTS proved highly tolerant of non-protic molecules (e.g. alkenes, alkynes, tertiary amines, ethers, heteroaromatics) but also of some protic groups (e.g. anilines, tertiary alcohols, secondary amine, indole, pyrrole, ureas, or even benzoic acid). On the other hand, primary and secondary alcohols, gentisic acid and DL-α-Tocopherol were highly detrimental. Water had no impact up to 2 000 ppm whilst higher contents decreased RCCs sharply. TEMPO, Oxo-TEMPO or BHT as putative radioprotectants were well tolerated.

Conclusion: The adapted HTS protocol proved efficient and reproducible, allowing us to expedite the entire 174 experiments in just four days of lab work. Radiofluorination with DBTS proved to tolerate a wide range of functional groups, notably including some protic functional groups. Here we demonstrated the versatility of DBTS and that systematic protection of protic groups may not always be always necessary, thus greatly simplifying both the organic synthesis of precursors and the radiolabeling process.

[1] J. Am. Chem. Soc. 2018, 140, 11125–11132.

[2] N. Engl. J. Med. 2025, 393, 2168–2170.

[3] a) J. Am. Chem. Soc. 2017, 139, 8267–8276. b) J. Am. Chem. Soc. 2024, 146, 10581–10590.