Organic azides are valuable intermediates in organic synthesis and as a result various methods have been developed for their synthesis.^[1] Traditionally, α-azidoketones are prepared by direct nucleophilic substitution of α-halo ketone with azide ions.^[2] Alternative methods are based on the reaction of an alkene or enol ether with an oxidizing agent in the presence of an azide source (Scheme 1).^[3] However, most of these methods still require a large excess of expensive and hazardous oxidants, which would increase cost and environmental concerns, and limit their practical utility. From a green and sustainable points of view, the direct utilization of O₂ as both oxidant and oxygen source represents an ideal methodology for the construction of oxygen-containing compounds. In 2018, Yang and co-workers^[4] reported a facile visible-light enabled method for the construction of α-azidoketones via metal- free Rose Bengal and (PhSe)₂ co-catalyzed oxyazidation of styrene with molecular oxygen. In 2021, Zhang and co- workers^[5] reported an electrochemical synthetic route to access α-azido ketones from styrene with molecular oxygen. Although these two methods are relatively mild, the reaction substrates primarily focus on styrene derivatives and do not extend to alkyl-substituted olefins. Therefore, the development of a simple, efficient, and environmentally friendly method for the construction of various α-azido- ketones is still highly desirable.

Alkenyl silanes are highly attractive intermediates due to their versatility in organic synthesis and materials science.^[6] Since the presence of 3d empty orbitals of silicon atoms has the α-effect on stable α-C radical, alkenyl silanes can be regarded as effective radical acceptor.^[7] In 1993, Isoe and co-workers^[8] reported an electron-promoted aerobic oxythiolation of alkenyl silanes with PhSH to deliver the α-phenylthiol ketone. In 2003, Oshima and co-workers^[9] reported that an Et₃B-promoted radical addition followed aerobic oxidation of alkenyl silanes with alkyl iodides to afford ketones. Although the oxygenation of alkenyl silanes to obtain the corresponding α-phenylthio carbonyl compounds and ketones has been reported, using alkenyl silanes as substrates to construct C—N bond has not been reported. Based on our previous efforts in the construction of the C—N bond and aerobic oxidation reactions,^[10] herein, we report a visible light-promoted aerobic oxidation of alkenyl silanes with TMSN₃ to afford α-azido- ketones.

At the beginning of our study, hex-1-en-2-yldiphenyl- silane (1a) was used as a model substrate, TMSN₃ as a nitrogen source, Acr^＋$\mathrm{MesClO}_{4}^{-}$ as a photocatalyst and MeCN as a solvent under the irradiation of 8 W blue LEDs for 18 h in the presence of an O₂ balloon. Initially, the reaction was conducted, delivering the final product in 71% yield (Table 1, Entry 1). However, screening a series of photocatalysts, the reactions afforded 3a in 13%~53% yields (Entries 2~5). After the optimization of the solvents, no better results were obtained (Entries 6~8). The yield was slightly reduced to 63% when the loading of Acr^＋$\mathrm{MesClO}_{4}^{-}$ was decreased to 2 mol% (Entry 9). Although the yield of 3a was increased to 80% with 10 mol% Acr^＋$\mathrm{MesClO}_{4}^{-}$, 5 mol% Acr^＋$\mathrm{MesClO}_{4}^{-}$ was chosen to further extend the application of this reaction (Entry 10).

Under the optimized conditions, the substrate scope was explored as shown in Table 2. The linear aliphatic vinyl-silanes could be reacted to afford 3a~3e in 50%~72% yields. (1-Cyclohexylvinyl)diphenylsilane could be converted to the desired products 3f in 54% yield. It is worth noting that internal vinylsilanes could react smoothly to deliver the products 3g~3h in 51%~54% yields. The diphenyl(1-phenylvinyl)silane with a methyl group on para-, ortho-positions could be tolerated giving the corresponding products 3i~3j in 38%~42% yields. The methoxy group at para-position on aryl rings displayed moderate reactivity to afford the targeted products 3k in 54% yield. Moreover, the reactions of disubstituted styryl silanes and heterocyclic silanes bearing 3-chloro-4-methyl- phenyl, 3-thienyl and benzo[b]thiophen-5-yl could react with TMSN₃ smoothly to produce the products 3l~3n in 37%~46% yields. The substrate containing bioactive molecule such as niflumic acid 3o could be also successfully accommodated, further demonstrating the utility of the approach. Furthermore, the various silane groups, such as dimethylphenyl silyl and diphenylmethyl silyl were found to be suitable for this transformation.

The reaction in the presence of 2,2,6,6-tetramethyl-piperidinooxy (TEMPO) afforded a trace amount of 3a with 98% recovery of 1a (Scheme 2a), which indicated that the reaction likely proceeds through a radical process. In addition, to further explore the scalability of this protocol, the gram-scale reactions were conducted by reacting 1a and 1k with 2 to afford targeted compounds in 55% and 64% yields (Scheme 2b), respectively. To explore the diversity of the reaction types, various radical sources, including trifluoromethyl radical, thiocyanate radical, bro- mide radical, chlorine radical, participated well in the reaction, affording the carbonyl compounds 4~7 in 54%~68% yields (Scheme 2c).

The plausible mechanisms for this transformation are proposed in Scheme 3. The photosensitizer (PC^＋) absorbs visible light under the irradiation of 8W blue LEDs to generate its excited state (^*PC^＋) which could oxidize TMSN₃ 2 to the azido radical A, which is then trapped by alkenyl silanes 1 to form B and the reduced species (PC^*).^[11] Sub-sequently, PC^* could be oxidized by dioxygen to provide superoxide and regenerate the photocatalyst (PC^＋). The radical B which then reacts with superoxide and TMS^＋ to afford the peroxide C. Then peroxide C is then eventually converted into the carbonyl product through migration of the silyl group to the internal oxygen atom. Finally, desiliconization of intermediate D produces the desired product 3.^[12]

In summary, a convenient method for the construction of α-azidoketones has been developed. The conditions are mild, and a variety of alkenyl silanes participate efficiently in the reaction. The reaction could be easily carried out in gram scale. A variety of radical sources, including trifluoromethyl radical, thiocyanate radical, bromide radical, chlorine radical, were all effectively applicable in the sustainable reaction system. The products are easily derivatized further. Current efforts in our laboratory are underway to explore various reactions of alkenyl silanes.