Photoredox-catalyzed aminofluorosulfonylation of unactivated olefins†

The development of efficient approaches to access sulfonyl fluorides is of great significance because of the widespread applications of these structural motifs in many areas, among which the emerging sulfur(vi) fluoride exchange (SuFEx) click chemistry is the most prominent. Here, we report the first three-component aminofluorosulfonylation of unactivated olefins by merging photoredox-catalyzed proton-coupled electron transfer (PCET) activation with radical relay processes. Various aliphatic sulfonyl fluorides featuring a privileged 5-membered heterocyclic core have been efficiently afforded under mild conditions with good functional group tolerance. The synthetic potential of the sulfonyl fluoride products has been examined by diverse transformations including SuFEx reactions and transition metal-catalyzed cross-coupling reactions. Mechanistic studies demonstrate that amidyl radicals, alkyl radicals and sulfonyl radicals are involved in this difunctionalization transformation.


Introduction
The sulfur(VI) uoride exchange (SuFEx) reaction revived by Sharpless and co-workers in 2014 is an emerging and promising click reaction that rests on the unique reactivity-stability balance of higher organosulfur uorides. 1 Sulfonyl uorides, some of the most widely used connective hubs for SuFEx click chemistry, have attracted enormous attention and nd widespread applications in elds as diverse as organic synthesis, 2 materials science, 3 chemical biology and drug discovery. 4 As a result, the development of efficient approaches for preparing sulfonyl uorides is undoubtedly in high demand and has become of special interest in synthetic chemistry. 5 However, compared to the tremendous progress made in the synthesis of aryl sulfonyl uorides, methods for accessing aliphatic sulfonyl uorides remain less explored. Conventionally, aliphatic sulfonyl uorides are prepared via uoride-chloride exchange of corresponding sulfonyl chlorides with uoride salts. 1 Alternatively, conversion of alkyl halides, thiols, or sultones into aliphatic sulfonyl uorides has also been achieved through multistep sequences. 6 Moreover, ethene sulfonyl uoride (ESF) has been used as a versatile building block for the synthesis of ethyl sulfonyl uoride derivatives. 5a,b,7 Despite these signicant advances, the development of more efficient methods to access aliphatic sulfonyl uorides is of high interest, because many pharmaceutical agents contain these structural motifs (Fig. 1a).
The direct difunctionalization of alkenes is a powerful strategy for the rapid assembly of molecular complexity and diversity. 8 With our continuous research interest in sulfonyl uoride synthesis, 5a,9 we intended to achieve the radical 1,2difunctionalization of unactivated alkenes providing functionalized aliphatic sulfonyl uoride derivatives. To the best of our knowledge, the sole example so far is uoroalkylation- uorosulfonylation of alkenes recently developed by the group of Liu and Chen. 10 However, a stoichiometric amount of metal reagent, such as a silver salt or zinc powder, was required to mediate these processes. Therefore, further endeavors to develop redox-neutral uorosulfonylation involving difunctionalization of alkenes to enrich the structural diversity of the sulfonyl uoride molecules are highly valuable.
b-Amino-substituted sulfonyl uorides are unique structural motifs with biologically important activities in various pharmaceuticals, in particular the peptide-type covalent inhibitors as illustrated in Fig. 1a. 4c-f Typically, these compounds were prepared from a-amino acids in a multi-step manner (Fig. 1b). 4d-f Inspired by the signicant progress in visible-light photoredox-catalyzed 1,2-difunctionalization of alkenes, 11 we envisioned that the radical aminouorosulfonylation might directly provide valuable b-amino sulfonyl uoride derivatives. Recently, Knowles and co-workers reported the generation of amidyl radicals through photocatalytic proton-coupled electron transfer (PCET) activation of amides, 12 and various transformations for difunctionalization of alkenes, including aminoalkylation, 13 hydroamination, 14 aminoalkynylation, 15 aminoarylation 16 and aminoacylation, 17 were elegantly realized beneting from the rapid C-centered radical formation through 5-exo-trig cyclization of the amidyl radical (k ¼ $10 5 s À1 ). Based on our previous experience on aryl sulfonyl uoride synthesis, 9 we questioned whether the alkyl radical could be sequentially trapped through SO 2 insertion and subsequent uorine transfer, enabling the introduction of both amino and uorosulfonyl groups across alkenes to access b-amino-substituted sulfonyl uorides (Fig. 1c).
However, in this process several challenges remain to be addressed: (1) a relatively low thermodynamic driving force for the conversion of amidyl into 1 or 2 alkyl radicals (DG 0 z À3 to À5 kcal mol À1 ) was unfavorable for the difunctionalization process; 16,18 (2) severe competitive reactions such as hydroamination and aminouorination might be observed; 14,19 (3) potential incompatibility of photocatalytic conditions with redox-active SO 2 and uorine sources. With these challenges in mind, herein we set out to describe a three-component ami-nouorosulfonylation of unactivated alkenes by merging photocatalytic PCET activation with a radical relay process.

Results and discussion
Initially, we conducted an optimization study using N-phenyl pent-4-enamide (1a) as the model substrate and it is readily accessible from aniline and 4-pentenoic acid. Gratifyingly, when 1a was treated with DABSO and NFSI in CH 3 CN in the presence of [Ir(dF(CF 3 )ppy) 2 (bpy)]PF 6 (PC-II, E 1/2 (*Ir III /Ir II ) ¼ +1.32 V vs. SCE) 20 and K 3 PO 4 under irradiation with blue LEDs for 10 hours, the desired aminouorosulfonylation product was smoothly obtained in 64% 19 F NMR yield ( Table 1, entry 1). When the photocatalyst was switched from PC-II to others, such as Ir-based photocatalysts [Ir(dF(CF 3 )ppy) 2 (dtbbpy)]PF 6 (PC-I) 21 and [Ir(dF(CF 3 )ppy) 2 (5,5 0 -dCF 3 bpy)]PF 6 (PC-III), 14a,22 4CzIPN, 23 Eosin Y, 24 and [Ru(bpy) 3 ]Cl 2 , 25 the yields decreased (entries 2-6). Substituting DABSO with other surrogates of sulfur dioxide (such as Na 2 S 2 O 5 and Rongalite), 26 or replacing the uorine donor NFSI with Selectuor led to a signicantly lower conversion or no reaction (entries 7-9). Screening of the bases revealed that K 3 PO 4 was the optimal choice, while using other inorganic or organic bases resulted in diminished yields (entries 10 and 11). Moreover, control experiments revealed that a photocatalyst and light irradiation were essential for the success of this transformation (entries 12 and 13). In the absence of a base, a lower yield was obtained (entry 14). For full details of the reaction optimization, see the ESI. † With the optimized conditions in hand, we next explored the substrate scope of the aminouorosulfonylation reactions, and the results are summarized in Schemes 1-3. To our delight, good yields were obtained for a wide range of anilide derivatives bearing different substituents on the arylamine moiety (2a-o). Various substrates bearing either electron-withdrawing or electron-donating substituents at the para position of the N-aryl groups were tolerated under the reaction conditions, furnishing pyrrolidinone-derived sulfonyl uorides 2a-i in moderate to good yields. However, lower yields were obtained for substrates with a strongly electron-withdrawing substituent such as CF 3 . The reactions of ortho-, meta-or di-substituted N-aryl amides also proceeded smoothly (2j-n). Notably, N-heteroaryl amides proved to be competent substrates in this transformation as well (2p, 2q).
With respect to the olen component, a variety of olens with different substituent patterns were successfully adapted (2r-z). As for terminal olens, substrates bearing various substituents at the a-carbonyl position, such as methyl, dimethyl, aryl, benzyl, and even bulky protected amino groups, were generally compatible in the reaction and induced moderate diastereoselectivities (Scheme 2, 2r-x). Nonterminal olen substrates were also well tolerated to deliver the corresponding secondary alkyl sulfonyl uorides (2y-z) in good yields. Remarkably, substrates bearing an endocyclic double bond could also be applicable for providing more complex fused polycyclic structures (2aa-ad) with excellent diastereoselectivities in some cases. It should also be mentioned that NFSI was fully consumed in most of these reactions, and phenylsulfonyl uoride was obtained as the side product, 27 which led to incomplete conversion of the amide substrates.
In addition to amide substrates, the amino-uorosulfonylation of carbamates and ureas was also examined under the standard conditions. Acyclic carbamates derived from substituted allylic alcohols could also undergo a cyclization cascade to provide access to sulfonyl uorides with oxazolidinone backbones (Scheme 3, 2ae-ai). A cyclohexenolderivatized carbamate could also be utilized to afford fused bicyclic product 2aj in good yield and diastereoselectivity. Similarly, b,g-unsaturated aryl urea could be employed in this transformation to assemble the imidazolidinone scaffold (2ak), even though a lower yield was obtained. Furthermore, the potential of this reaction was evaluated with more challenging substrates derived from pharmaceuticals and natural products. The amides derived from sulfamethazine (antibacterial) and lenalidomide (anticancer) were successfully cyclized to deliver sulfonyl uoride products 2al and 2am in moderate yields. Similarly, menthol and estrone derivatives were well tolerated to provide the desired products 2an and 2ao, respectively.
With success in the preparation of 2a on a 1 mmol scale without noticeable erosion in yield (58%), we then investigated diversication of 2a through a wide variety of SuFEx click reactions (Scheme 4). As demonstrated in Scheme 4, pyrrolidinone-based sulfonyl uoride 2a readily underwent SuFEx with methanol, phenols, TBS-protected mecarbinate and TMS-protected diacetonefructose, affording the corresponding Scheme 1 Scope of N-(hetero)aryl amides. Reaction conditions as stated in Table 1, entry 1. Table 1, entry 1. Diastereomeric ratios were determined by NMR analysis of the crude reaction mixtures.

Scheme 2 Scope of terminal and nonterminal olefins. Reaction conditions as stated in
sulfonate esters D1-D4 in the presence of a base or silicon additives. Likewise, S(VI)-N bonds were smoothly formed to give sulfonamides D5, D6, and sulfonyl azide D7, and D7 could be further transformed into sulfonyl triazole D8 via a coppercatalyzed azide-alkyne click reaction.
The synthetic utility of sulfonyl uorides was also demonstrated by cross-coupling reactions (Scheme 5). As shown in Scheme 5, 2d could be used as the coupling partner in Pdcatalyzed Suzuki and Sonogashira reactions, which proceeded chemoselectively at the para-bromophenyl moiety of 2d affording D9 and D10 with 30% and 20% yields (not optimized), respectively. Additionally, the pyrrolidinone skeleton of 2a could be smoothly reduced to pyrrolidine D11 with 9-borabicyclo [3.3.1]nonane . Taken together, the abovementioned transformations demonstrated the chemical stability and robustness of alkyl sulfonyl uorides, and also broaden their applications in organic synthesis.
To gain mechanistic insight into this three-component aminouorosulfonylation reaction, several control experiments were carried out (Scheme 6). First, the aminouorination products could be detected in some cases (less than 5% yield). In contrast, when the reaction was performed in the absence of DABSO, the uorinated product could be isolated in up to 43% yields (Scheme 6a). Then, the formation of 2c was almost completely suppressed when a radical scavenger TEMPO (2,2,6,6-tetramethyl-1-piperidinyloxy) was added to the reaction, and the trapping product could be detected by LC-MS (Scheme 6b, see the ESI † for details). Next, a trace amount of 2c was detected when a milder radical scavenger 1,1-diphenylethylene was introduced into this reaction, and the olenation products D12 and D13 were detected by LC-MS, which suggested that an amidyl radical might be generated (see the ESI † for details). Meanwhile, the sulfur dioxide insertion product D14 was Scheme 3 Scope of carbamates, ureas, pharmaceuticals and natural products. Reaction conditions as stated in Table 1, entry 1. Diastereomeric ratios were determined by NMR analysis of the crude reaction mixtures. isolated in 33% yield (Scheme 6c), indicating the existence of an alkyl sulfonyl radical in this transformation. Furthermore, the observation of an aza-Michael product with ethyl acrylate suggested that the amidyl anion is formed under these conditions 28 (Scheme 6d). Finally, Stern-Volmer studies showed that the potassium salt of 1a could quench the excited Ir photocatalyst (Scheme 6e, see the ESI † for details).
On the basis of these mechanistic experiments and related literature reports, [12][13][14][15][16][17]28 we propose a mechanistic scenario initiated by the formation of an amidyl radical A through a stepwise or concerted proton-coupled electron transfer (PCET) process (Scheme 7). Subsequent intramolecular addition to the unactivated olen results in the formation of a g-lactam-bearing alkyl radical B. Then, trapping of the alkyl radical B with SO 2 affords an alkylsulfonyl radical C. Subsequent uorine atom transfer from NFSI provides the sulfonyl uoride product. Meanwhile, the (PhSO 2 ) 2 N radical generated from N-uorobenzenesulfonimide (NFSI) (E pc ¼ À0.78 V vs. SCE in MeCN) 29 could accept one electron from [Ir II ] to regenerate the photocatalyst (E 1/2 (Ir III /Ir II ) ¼ À1.37 V vs. SCE). 30

Conclusions
In conclusion, the rst three-component amino-uorosulfonylation of unactivated alkenes has been developed for the synthesis of sulfonyl uorides by merging photocatalytic proton-coupled electron transfer (PCET) with radical relay processes. Diverse aliphatic sulfonyl uorides featuring medicinally privileged heterocyclic scaffolds (pyrrolidinone, oxazolidinone and imidazolidinone) have been efficiently provided under mild conditions, employing easy-to-handle DABSO and NFSI as the sulfur dioxide surrogate and uorine source, respectively. The SO 2 F-containing products obtained could be used for further diversication through SuFEx click reactions and transition metal-catalyzed cross-coupling reactions. Control experiments and Stern-Volmer studies have revealed that a PCET-based activation is key to the formation of amidyl radicals and subsequent alkyl and sulfonyl radicals. Further elaboration of this difunctionalization strategy for the synthesis of structurally diverse sulfonyl uorides towards biological applications is ongoing in our laboratory.

Data availability
The electronic supplementary information include experimental detail, NMR data and HRMS data.

Conflicts of interest
There are no conicts to declare.