Reaction development. Initially, we employed methyl 3-(p-tolylsulfinyl)propanoate 1a, toluene 2a as the model substrates to optimize the conditions of the TM-free C–H thiolation reaction. After extensive screening of various reaction parameters, the optimal conditions were achieved to be 1a (1.3 equiv), 2a (0.3 mmol), Tf2O (1.3 equiv), solvent (3 mL) under nitrogen atmosphere at room temperature for 1 h, then Et3N (5 equiv) at room temperature for 1 h (Table 1 entry 1). A series of bases were screened, and no better yields were obtained than Et3N (Table 1 entries 2–5). Switching the solvent to CH3CN only provide the desired product in low yield (Table S1, entry 6). The yield was reduced to 30% when TFAA was used in lieu of Tf2O (entry 7). A similar result was obtained using ethyl ester 1b as the thiolation reagent (entry 8). However, tert-butyl ester 1c only delivered the desired product in 23% yield (entry 9).
With the established optimal reaction conditions in hand, we first investigated the substrate scope of arenes (Table 2). A wide range of substituted benzenes could be converted to the corresponding diaryl sulfides in good yields (3a–s). Functional groups including cyclopropyl, halogens, allyloxy, and ester were well tolerated in this transformation. It is noteworthy that the reactivity of this thiolation reaction was not affected by steric hindrance of the arene substrates (3q and 3s). To our delight, heteroaromatics could also be applied to the reaction system without obstacle (3t-w). Among them, indole, pyrrole, and thiophene could give the corresponding products in the excellent yields. Although quinoline derivative was less reactive under standard conditions, increasing reaction temperature slightly improved product yield (3u). Notably, this strategy was successfully applied to the late-stage C–H thiolation of a variety of drug-like molecules (3aa–af). (S)-4-benzyl-2-oxazolidinone derivative, nimesulide, bifonazole, L-phenylalanine derivative, estrone derivative and D-salicin derivative all could be modified by this method. We then examined the reaction of 1,2-dimethoxybenzene 2o with a range of β-sulfinylesters. Aryl sulfoxides bearing electron-donating groups or halogens had good reactivity (4a–f), whereas aryl sulfoxide with a nitro group only delivered the desired product in 26% yield (4g). In addition, pyridyl and thienyl sulfoxides were compatible, albeit the desired products were formed in low yields (4h and 4i). Remarkably, alkyl sulfoxides including benzyl sulfoxide were applicable to the reaction system, giving the desired products in good yields (4j–l).36
Inspired by these exciting results, we discussed the application of this strategy in the synthesis of more challenging alkenyl, alkynyl, and alkyl sulfides (Table 4). Styrenes containing electron-donating and -withdrawing groups all have good adaptability (6a–j). It is worth mentioning that high E-selectivities were obtained in all cases. Vinyl sulfide 6k and allyl sulfide 6k` were formed in 93% combined yield using α-methylstyrene as substrate. 1,1-Diphenylethene and triphenylethene were smoothly converted to the desired alkenyl sulfides in 89% and 46% yields, respectively (6l and 6m). When tetraphenylethene was subjected to the reaction system, only aryl C–H thiolation product was observed (6n). It should be noted that compounds (6l–n) were promising aggregation-induced emission luminogens.53 Additionally, a range of aryl and alkyl sulfoxides could also donate the corresponding alkenyl sulfides without obstacle (6o–y).
Next, we proceed to investigate the application of this strategy to C(sp)–H thiolation. Gratifyingly, the reaction was proved to be applicable to C(sp)–H thiolation of aryl alkynes (8a–h). Low reactivity of alkyl alkynes was observed, and only 17% yield was obtained (8i). In addition to thioarylation (8j–p), thioalkylation of C(sp)–H bond was also successful (8q and 8r).
Finally, we found that 1,3-dicarbonyl compounds could be thiolated with β-sulfinylesters to give a variety of alkyl sulfides in 38-88% yields (10a–i).
We have carried out a gram-scale reaction of dimethyl 3,3'-sulfinyldipropionate 1m and p-xylene 5m, giving the sulfide 11 in 74% yield (Fig. 2a). We then sought to investigate the possibility of performing iterative C–H thiolation, wherein the newly synthesized sulfoxide 12 may serve as the starting material for an additional C–H thiolation. Delightfully, sulfides 13 and 14 were formed in 72% and 73% yields from 4-bromostyrene 5i and D-salicin derivative 2af, respectively (Fig. 2b). Finally, the proposed methyl acrylate byproduct could be captured via Heck reaction to give 15 and 16 in 25% and 40% yields, respectively (Fig. 2c).