Introduction

C-Sn bond is known as a kind of very important carbon-metal bond, which has numerous applications in organic chemistry1,2, biological chemistry3,4,5, and medicinal chemistry6,7. There are three main routes for the transformation of organostannanes, including transmetalation process8,9,10, carbanion process11, and radical process12,13. There are lots of reports on transmetalation process. For example, the Stille cross-coupling which is a powerful tool for the formation of C-C bond has been widely applied in synthetically organic chemistry (Fig. 1a)8,9,10. The C-Sn bond transformation via carbanion process has also been well studied. For instance, Komatsu et al. reported that a kind of aza five-membered heterocycles could be efficiently synthesized from N-(stannylmethyl)thioamides via a ylide key intermediate (Fig. 1b)11. However, achievements via radical process are rare. In 1995, Baciocchi et al. reported the Fe-catalyzed oxidation of benzyltrialkylstannanes by iodosylbenzene, which proceeded via a radical key intermediate. However, it showed very low value in synthetic chemistry, since only a complicated mixture could be afforded as the final product (Fig. 1c)12. Obviously, research in this field remained undeveloped. We wondered whether there is a new method that can efficiently generate highly reactive carbon radical intermediate through C-Sn bond cleavage, and more importantly, the subsequent transformation of the in situ generated carbon radical intermediate should be highly selective to make such a reaction synthetically useful. In our previous work, we reported the photoinduced HBr-catalyzed C-Si bond cleavage of benzylsilanes, in which benzyl radicals could be generated and highly selectively oxidized into benzoic acids14. In this reaction, the bromine radical generated from the pre-catalyst HBr could abstract a single electron from the C-Si σ bond, which would finally break the C-Si bond and form a benzyl radical. Considering the similarity of C-Sn and C-Si bond, we assumed that benzylstannane might also be converted into benzyl radical under similar reaction conditions. If it works, we will be able to develop a new method for the transformation of organostannanes via radical process. Herein, we wish to report our recent observation in this field (Fig. 1d).

Figure 1
figure 1

Transformation of organostannanes.

Full size image

Results and Discussion

Optimization and scope investigation

We conducted a series of reactions to screen the reaction conditions. Initially, the solvent effect was explored carefully. A solution of (4-methoxybenzyl)- trimethylstannane 1a (0.2 mmol) and 20 mol% of HBr (aq., 48%) in different solvents (10 mL) was irradiated by a 300 W Xe lamp at rt under air atmosphere (entries 1–5, Table 1). A mixture of 4-methoxybenzaldehyde 2a and 4-methoxybenzoic acid 3a in a very poor selectivity was afforded when CH3CN or dimethyl carbonate (DMC) was used as the solvent (entries 1 and 2, Table 1). When dichloromethane (DCM) or acetone was used as the solvent, 3a was formed as the major product in a better yield (entries 3 and 4, Table 1). To our delight, the utilization of ethyl acetate (EA) as the solvent delivered 3a as the sole product in an 89% isolated yield (entry 5, Table 1). Thus, EA was chosen as the best solvent. Next, some bromide salts of group I elements, like LiBr, NaBr, or KBr, were tested, however, all the results were not satisfactory (entries 6–8, Table 1). Decreasing the amount of HBr to 10 mol% led to a diminished yield and a much lower chemoselectivity (entry 9, Table 1). When this reaction was run in the absence of light, no product was generated (entry 10, Table 1), which indicated that light played an important role in the reaction process. Thus, Condition A (1a

Table 1 Optimization of reaction conditionsa.
Full size table

(0.2 mmol), 20 mol% of HBr (aq., 48%), EA (10 mL), air (1 atm), 300 W Xe lamp, quartz, and rt) was chosen as the optimized reaction condition for the following studies.

With the optimized reaction conditions in hand, we investigated the substrate scope of this reaction with a series of benzyltrimethylstannane derivatives. As shown in Table 2, the reaction showed good tolerance. Substrates with electron-donating groups were found to be reactive, giving the corresponding products in good yields (3a–e). The substrate with no substitution on the phenyl ring, like benzyltrimethylstannane 1f, was converted into 3 f in an excellent yield. Nice yields were also obtained when electron-withdrawing groups were introduced. For substrate with a weak electron-withdrawing group, like chlorine or fluorine atom, the corresponding product was isolated in good yields (3 g and 3 h). While the yields were a little lower, when the substitutions were changed into strong electron-withdrawing groups (3i–l). We were delight to see that naphthyl was also tolerant in this reaction, affording the corresponding 1- or 2-naphthoic acid, respectively (3 m and 3n). Importantly, the benzoic acid derivatives 3 were formed as the sole product in all the above tests.

Table 2 Photooxidation of 1a–n under Condition Aa.
Full size table

Interestingly, when trimethyl(phenethyl)stannane 1o was applied under Condition A, only benzoic acid 3f was formed in a good yield (Fig. 2). We assumed that phenylethyl radical was generated and it would be further oxidized into benzoic acid 3f as the final product.

Figure 2
figure 2

Photooxidation of 1o under Condition A.

Full size image

In addition to the above primary alkyl stannanes, the reactivity of secondary alkyl stannanes were also examined. The corresponding ketones were formed as the sole product under Condition A (Fig. 3).

Figure 3
figure 3

Photooxidation of 1p and 1q under Condition A.

Full size image

Mechanism studies

In order to get a better understanding of this reaction, a series of control experiments were conducted to investigate the mechanism. As shown in Table 3, the photo reaction of 1a proceeded smoothly to afford 3a in a 94% NMR yield under Condition A (entry 1, Table 3). When the reaction was carried out without light, no reaction occurred and 91% of 1a was recovered (entry 2, Table 3), which meant that light was the essential condition to start this reaction. When this reaction was run in the absence of air, 89% of 1a was recovered without any other products isolated (entry 3, Table 3), which indicated that this reaction did not proceed without air. Next, the reaction was run for only 3 h (entry 4, Table 3). In this case, 1a was not fully consumed. The corresponding benzaldehyde 2a was formed. The desired benzoic acid 3a was afforded in a low yield. This result might be a proof that benzaldehyde should be a key intermediate in the whole transformation. Then another reaction was run for 3 h under photo irradiation and a following 3 h without light (entry 5, Table 3). Almost the same result as the above test (entry 4, Table 3) was observed, which indicated that light played an important role not only in the initiation period, but also throughout the whole reaction process. Finally, the reaction was conducted without the catalyst HBr (entry 6, Table 3). The reaction speed was dramatically decreased with 46% of 1a recovered. Only 9% of 3a was yielded. Meanwhile, 46% of 2a was formed. These results clearly showed that HBr not only facilitated the C-Sn bond cleavage at the beginning of reaction, but also benefited in the following oxidation step.

Table 3 Photo reaction of 1a under different conditionsa.
Full size table

To further prove that benzaldehyde was the intermediate of this whole transformation, 2a was directly employed under Condition A. After 4 h, 3a was afforded in a 99% yield, which fully supported this inference (Fig. 4).

Figure 4
figure 4

Photo reaction of 2a under Condition A.

Full size image

According to our previous work14, Br2 was formed under similar photo reaction conditions and played a key role in the catalytic process. Thus, 10 mol% of Br2 was employed instead of HBr to identify its catalytic property in this reaction. As a result, 2a was formed in a 51% yield and 3a was afforded in a 36% yield after irradiated for 6 h (Fig. 5), which indicated that Br2 could also catalyze this reaction with relatively lower efficiency.

Figure 5
figure 5

Br2-catalyzed photo reaction of 1a.

Full size image

Based on the above experiment results and literature precedents14,15,16,17,18,19,20, a possible mechanism was proposed as shown in Fig. 6. Firstly, Br− was oxidized into Br2 by oxygen under photo irradiation14. Br2 was then photochemically converted into bromine radical15,16,21,22. In the presence of HBr, bromine radical had a relatively high oxidation potential18,23 which was strong enough to capture one electron from C-Sn σ bond of 1. Thus, the transformation of organostannanes was initiated. After being grabbed one electron from the C-Sn bond, the highly unstable radical cation 4 was formed. The subsequent heterolytic cleavage of C-Sn bond in 4 proceeded quickly to give carbon radical 5. It could be easily oxidized into benzaldehyde 2 24,25,26, and further oxidized into the final product benzoic acid 3 27 under Condition A.

Figure 6
figure 6

Proposed mechanism.

Full size image

Conclusions

In summary, we have developed a novel method for the transformation of organostannanes via radical process. Under photo irradiation, HBr could efficiently catalyze the heterolytic cleavage of C-Sn bond in organostannanes and the subsequent highly selective oxidation. The mechanism was well studied. In this reaction, bromine acted as a medium of photochemical single electron transfer process. Such a catalyst is simple, effective, cheap, and easy to handle. Further studies in this field is in progress in our laboratory.

Methods

Synthesis of 4-methoxybenzoic acid (3a)

1a (57 mg, 0.20 mmol), EA (10 mL), and HBr (aq., 48%) (4.5 μL, 0.04 mmol) were added to a quartz reaction flask which was equipped with a magnetic stirrer and a condenser. The mixture was irradiated by a Xe lamp (300 W) at rt in the open air. The photoreaction was completed after 6 hours as monitored by TLC (eluent: petroleum ether: ethyl acetate = 10:1). The solvent was removed and the residue was purified by flash chromatography on silica gel (eluent: petroleum ether: ethyl acetate = 3:1) to afford 3a as a solid (27 mg, 89%); 1H NMR (400 MHz, CDCl3) δ 8.07 (d, J = 8.4 Hz, 2 H), 6.95 (d, J = 8.4 Hz, 2 H), 3.88 (s, 3 H).