On-Water Accelerated Dearomative Cycloaddition via Aquaphotocatalysis

The study addresses the challenge of effecting intermolecular visible-light [2+2] photocycloadditions of β-(hetero)arylated ethenesulfonyl fluorides (ESFs) with heteroaromatics, which in conventional organic solvents rapidly undergo E/Z isomerization and stall at equilibrium. Motivated by the unique kinetic and selectivity enhancements observed for reactions “on water,” the authors hypothesized that bulk water and the water–oil interface could accelerate photocycloadditions using water-insoluble substrates and hydrophobic photocatalysts. Given the prominence of SuFEx chemistry and the water stability of sulfonyl fluorides, creating heterocyclic alkyl SuFEx hubs via a simple, aqueous, photocatalytic protocol would expand applications in biomedicine and materials. The purpose of the study is to develop an on-water, visible-light aquaphotocatalytic dearomative [2+2] cycloaddition that overcomes solvent-induced isomerization, broadens substrate scope to biorelevant heteroaromatics, and enables downstream SuFEx ligations.

Water as a reaction medium can enhance rates and selectivities and is central to biological processes. The “on-water” effect—heightened reactivity at the water–oil interface for immiscible reactants—has been demonstrated for numerous catalytic systems (Lewis/Brønsted acids and bases, and transition metals), often for reactions with negative activation volume resembling high-pressure acceleration. Prior successes largely involve ionic mechanisms and Michael-type additions. However, significant rate acceleration for photocycloadditions in bulk aqueous media using water-insoluble substrates and hydrophobic photocatalysts has been lacking. SuFEx click chemistry couples electrophilic S(VI) centers with amines/alcohols, enabling bioconjugation to macromolecules such as proteins and DNA. Sulfonyl fluorides are water-stable and compatible with various catalyses; ESF is a modular unit accessible on kilogram scale, with β-substitution enabling further reactivity under water-accelerated organocatalysis. Yet visible-light photocycloadditions in organic solvents suffer from rapid E/Z isomerization, thwarting productive coupling. The authors build on these insights to test whether on-water conditions can uniquely enable such photocycloadditions.

Optimization: A biphasic, on-water photocatalytic system was evaluated for phenyl-ESF (1a) with 2,3-benzofuran (2a) under blue LED irradiation (50 W, 456 nm). Photocatalyst screening (2.0 mol%) in H2O (10 L/mol) revealed that Ru(bpy)3(PF6)2, Eosin Y (neutral or disodium), Rose Bengal, 9-mesityl-10-methylacridinium ClO4, and thioxanthone were inactive. Cyanoarenes (4DP-IPN) showed low conversion (14%). Ir catalysts improved yields (fac-Ir(ppy)3: 44%; [Ir{dF(CF3)ppy}2(dtbpy)]PF6: 77%). Organophotocatalyst 4Cz-IPN delivered >99% yield (exo:endo 79:21). Control experiments verified no reaction without light or catalyst; green (525 nm) or sunlight were inferior; 390 nm light was effective (86% yield). TEMPO suppressed the reaction and led to Z-1a'. Reaction media screening: Conventional organic solvents (acetone, CH2Cl2, DMF, DMSO, Et2O, MeCN, MeOH, n-hexane, PhMe, THF) failed, returning 38–96% 1a and 3–53% Z-1a'. Surfactant micellar media (TPGS-750-M or SPGS-550-M, 2 wt%) gave low conversion (12% and 23%), supporting the necessity of a water–oil interface. Neat conditions afforded 84% yield; D2O gave 90% vs H2O >99%; saturated NaCl(aq) was ineffective. Comparative kinetics showed faster initial rates on water vs neat. Substrate scope: Under optimized conditions (1: 0.4 mmol; 2: 2.0 mmol; 4Cz-IPN 2.0 mol%; H2O 10 L/mol; 24 h; vigorous stirring ~1150 rpm; 50 W 456 nm), diverse β-aryl ESFs (1a–1q) and heteroaromatics (2a–2l: benzofuran, benzothiophene, N-Boc indole, furan, substituted analogs, and a Boc-L-tryptophan derivative) furnished cyclobutanes in good to quantitative yields with diastereomeric mixtures (exo/endo assigned by NOESY and SC-XRD). β-unsubstituted ESF was inactive. Scale-up: Gram-scale run with 1a (1.00 g) and 2a (5 equiv) using 4Cz-IPN 0.5 mol% in H2O (10 L/mol), 48 h, gave 3aa in 95% combined yield (exo 74%, 1.21 g; endo 21%, 0.34 g); structures confirmed by X-ray (CCDC-2279317/2279318). SuFEx derivatizations: Post-cycloaddition transformations included formation of sulfonamide (benzylamine, 61%), sulfonyl azide (TMSN3/DMAP) followed by CuAAC to triazole (90% overall for click step), and sulfonate esters from phenols (86%). Bioactive substrates estrone and stanolone were converted to SuFEx adducts (64% and 31%). Mechanistic probes: Fluorescence quenching of 4Cz-IPN by 2a/1a was evaluated in THF/H2O mixtures; Stern–Volmer plots showed pronounced quenching as water fraction increased, correlating with partial heterogeneity (50:50 THF/H2O) and measurable product formation. Stirring-speed dependence showed higher conversion with increased rpm, implicating interfacial area. Using Z-1a' as starting material gave low conversion (14%) and diminished selectivity, indicating the productive pathway originates from E-1a. General procedure: β-arylated ESF (0.4 mmol, 1.0 equiv) and 4Cz-IPN (2.0 mol%) were combined with heteroarene (2.0 mmol, 5.0 equiv) and H2O (10 L/mol; 4.0 mL); vigorously stirred (>1000 rpm) under 50 W 456 nm blue LED for 24 h with cooling fan; worked up by EtOAc/brine extraction, drying, concentration, and column chromatography; recrystallization as needed.

• Water as the sole reaction medium enables an on-water accelerated dearomative [2+2] photocycloaddition between β-(hetero)arylated ESFs and heteroaromatics, whereas conventional organic solvents fail and induce E/Z isomerization.
• Optimal catalyst: 4Cz-IPN (2.0 mol%) delivered >99% conversion/yield with exo:endo = 79:21 (Table 1, entry 18). Ir catalysts were moderately effective (fac-Ir(ppy)3: 44%; [Ir{dF(CF3)ppy}2(dtbpy)]PF6: 77%). Many common photocatalysts were inactive.
• Controls: No reaction without light or catalyst; 390 nm light gave 86% yield; green light or sunlight were inferior; heating alone ineffective; TEMPO suppressed reaction and gave Z-1a' (46%).
• Media effects: Organic solvents returned 38–96% 1a and 3–53% Z-1a'. Surfactants reduced conversion (TPGS-750-M: 12%; SPGS-550-M: 23%), supporting interfacial on-water acceleration. Neat conditions gave 84% yield; D2O 90%; H2O >99%. Saturated NaCl(aq) ineffective.
• Kinetics: At 2 h, neat gave 38% vs on-water 80% of 3aa; at 12 h, neat 56% vs on-water 91%, indicating pronounced early-stage acceleration on water.
• Scope: Broad β-aryl ESFs (1a–1q) and heteroaromatics (2a–2l) afforded products in good-to-quantitative yields (many up to 98–99%). Diastereoselectivity generally favored exo; notable endo preference with 3-substituted heteroaromatics (e.g., 3ad exo:endo 15:85; 3ae 21:79; 3ag 44:56). β-Unsubstituted ESF was inactive.
• Scale-up: 1.00 g of 1a furnished 3aa in 95% combined yield (exo 74% [1.21 g], endo 21% [0.34 g]); structures confirmed by SC-XRD (CCDC-2279317/2279318).
• SuFEx utility: Post-functionalizations included sulfonamide (61%), sulfonyl azide then CuAAC to triazole (90%), and sulfonate ester (86%); bioactive molecules estrone (64%) and stanolone (31%) were successfully ligated.
• Mechanism: Data support a triplet energy transfer pathway with exciplex formation at the water–oil interface; stronger fluorescence quenching with higher water content; interfacial area (stirring speed) correlates with rate; Z-1a' as starting material is poorly productive (14%), indicating the E-isomer underpins productive [2+2].

The findings demonstrate that bulk aqueous media uniquely enable a visible-light [2+2] cycloaddition of β-arylated ESFs with heteroaromatics by leveraging the water–oil interface to accelerate productive triplet energy transfer and exciplex formation. In organic solvents, rapid E/Z isomerization dominates, preventing cycloaddition; on water, high-pressure-like interfacial effects and hydrophobic confinement favor encounter and coupling over isomerization, delivering high yields and diastereoselectivity. Broad substrate scope, gram-scale feasibility, and downstream SuFEx ligations with biologically relevant partners showcase the method’s practical relevance for constructing heterocyclic alkyl SuFEx hubs. Mechanistic probes (fluorescence quenching trends with water fraction, stirring-speed dependence, low productivity of Z-1a', and TEMPO inhibition) collectively support an interfacial, triplet EnT mechanism with exciplex intermediacy, rationalizing the observed on-water acceleration and product selectivity.

An on-water accelerated dearomative aquaphotocatalysis has been developed for accessing heterocyclic alkyl SuFEx hubs via visible-light [2+2] cycloaddition between β-arylated ethenesulfonyl fluorides and heteroaromatics. Water is essential, transforming an otherwise isomerization-prone process in organic solvents into a high-yielding reaction with broad scope, preparative scalability, and immediate utility in SuFEx ligations to form sulfonamides, sulfonyl azides/triazoles, and sulfonate esters, including with bioactive molecules. Mechanistic evidence supports a triplet energy transfer pathway enhanced at the water–oil interface via exciplex formation and hydrophobic confinement. Future work could extend this platform to additional heteroarene classes, enable enantioselective variants, integrate with DNA-encoded library and antibody–drug conjugate syntheses under aqueous conditions, and further quantify interfacial dynamics governing rate and selectivity.

• β-Unsubstituted ESF substrates were inactive under the developed conditions.
• The reaction requires bulk water and vigorous stirring; conventional organic solvents fail, and surfactant micelles reduce conversion, indicating reliance on heterogeneous interfacial conditions.
• Diastereomeric mixtures (exo/endo) are generally formed, requiring separation; endo selectivity predominates for some 3-substituted heteroaromatics without a clear mechanistic rationale.
• D2O reduces yield relative to H2O; saturated NaCl(aq) is ineffective.
• The Z-ESF isomer is a poor starting material (low conversion), and TEMPO inhibits the reaction, constraining conditions.
• Light-source sensitivity is noted (blue or purple effective; green/sunlight inferior), and typical reactions require 24 h irradiation; scale-up required 48 h.