Scheme 3 Trials to determine the allowable range of diaryliodonium salts.
a Reaction conditions:1a (0.20 mmol),2 (0.40 mmol), rongalite (0.80 mmol), Bu4NBr (1.0 equiv), DMSO (2.0 mL), 90 °C, 3 h, under Ar.b Isolated yields based on 1a .c X = OTf. d X = BF4.
Subsequently, the effects of substituents on the CF3-alkene were examined (Scheme 4). Using CF3-alkenes with alkyl (4a4c ), alkoxy (4d4f ), and halogen (4g4i ) substituents on the aromatic ring, the reaction was observed to proceed well, and the target products were isolated in yields ranging from 60% to 85%. The reaction involved bothortho - and meta -substituted CF3-alkenes, resulting in the excellent-yield products 4j and 4k , respectively. It was discovered that this reaction was also compatible with electron-withdrawing moieties such -CF3, -Ac, and -COOEt, which gave the corresponding products 4l4nin yields of 59–64%. Furthermore, heteroatom groups, such as trimethylsilyl (4o ), amine (4p ), -SMe (4q ) and -OH (4r ) moieties, were discovered to be suitable for this reaction. Experiments using polycyclic aromatic hydrocarbons as substituents established that products 4s4w could be obtained in good yields (65–82%). Also, this reaction worked well in the presence of heterocycle-substituted substrates (4x4ab ), with yields of 52−76%. In order to illustrate the usefulness of this reaction, the late-stage functionalizations of the anti-inflammatory drugs ibuprofen andnaproxen were assessed. These reactions gave the target products4ac and 4ad in yields of 73% and 66%, respectively.
Scheme 4 Trials to determine the allowable range of CF3-alkenes.
a Reaction conditions: 1 (0.20 mmol),2a (0.40 mmol), rongalite (0.80 mmol), Bu4NBr (1.0 equiv), DMSO (2.0 mL), 90 °C, 3 h, under Ar.b Isolated yields based on 1 .
In addition to the late-stage functionalizations stated above, this method was also carried out on a larger 5 mmol scale, yielding the target product in a 66% isolated yield (Scheme 5).
Scheme 5. Gram-scale synthesis. a For details, see the Supporting Information.
A number of control experiments were carried out to elucidate the mechanism underlying this reductive cross-coupling reaction. No significant formation of product 3a was observed in the case that CF3-alkene 1a was reacted with diaryliodonium salt 2a in the presence of TEMPO, indicating that this transformation likely involves radicals (Scheme 6a). A standard radical trapping experiment with 1,1-diphenylethylene was also carried out and radical adduct 5 was identified by gas chromatography-mass spectrometry while product 3a was obtained in a 58% yield (Scheme 6b). In other trials, 1.0 equiv of H2O was added to the reaction to trap any anions that may have been generated and the un-defluorinated product 3a′was detected by gas chromatography-mass spectrometry, while product3a was separated in a 72% yield (Scheme 6c). A plausible reaction mechanism is showed in Scheme 6d based on the present control experiments and previous work.[12] At the beginning, the pyrolysis of rongalite generates SO22- and also releases HCHO and H+. Subsequently, the SO22- that is gradually released and diaryliodonium salt 2 undergo a single electron transfer process to produce aryl radical A . The reaction of this radical and CF3-alkene 1 furnishes the radical intermediate B . Finally, B is reduced by either a sulfur dioxide anion or a sulfur dioxide radical anion to give anion species C . The latter undergoes defluoridation to deliver the desired gem -difluoroalkene product 3 . It is worth noting that the chemoselective radical reduction in this reaction is vital to realizing this transformation.
Scheme 6. Mechanistic study.
Conclusions
In summary, a transition-metal-free allylic defluorination reductive cross-coupling between CF3-alkenes and diaryliodonium salts was established for the construction ofgem -difluoroalkenes. The industrial product rongalite was employed as both a radical initiator and reductant. A catalyst was not required and the use of the control-release rongalite instead of a metal powder reducing agent promoted a sequential and highly selective single-electron transfer process. Through this method, anti-inflammatory drugs could be subjected to late-stage functionalization and a scaled-up version of this synthesis was also achieved.
Experimental
All the materials and solvents were commercially available and used without further purification. TLC analysis was performed using pre-coated glass plates. Column chromatography was performed using silica gel (200–300 mesh). 1H spectra were recorded in CDCl3 and DMSO-d 6 on 600/400 MHz NMR spectrometers and resonances (δ) are given in parts per million relative to tetramethylsilane. Data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, h = quintet, p = sextet, m = multiplet), coupling constants (Hz) and integration. 13C spectra were recorded in CDCl3 and DMSO-d 6 on 150/100 MHz NMR spectrometers and resonances (δ) are given in ppm.19F spectra were recorded in CDCl3 and DMSO-d 6 on 376 MHz NMR using TMS as internal standard. High-resolution mass spectra (HRMS) were obtained by electrospray ionization (ESI) on a TOF mass analyzer. The X-ray crystal-structure determinations of 3ae were obtained on a Bruker SMART APEX CCD system. Rongalite was commercially available (CAS No: 149-44-0) and purchased from TCI corporation.