Aerial Oxygen-Driven Selenocyclization of O-Vinylanilides Mediated by Coupled Fe3+/Fe2+ and I2/I− Redox Cycles

In the past decade, selenocyclization has been extensively exploited for the preparation of a wide range of selenylated heterocycles with versatile activities. Previously, selenium electrophile-based and FeCl3-promoted methods were employed for the synthesis of selenylated benzoxazines. However, these methods are limited by starting material availability and low atomic economy, respectively. Inspired by the recent catalytic selenocyclization approaches based on distinctive pathways, we rationally constructed an efficient and greener double-redox catalytic system for the access to diverse selenylated benzoxazines. The coupling of I2/I− and Fe3+/Fe2+ catalytic redox cycles enables aerial O2 to act as the driving force to promote the selenocyclization. Control and test redox experiments confirmed the roles of each component in the catalytic system, and a PhSeI-based pathway is proposed for the selenocyclization process.

As an important heterocyclic scaffold, 3,1-benzoxazine is widely found in natural products and bioactive molecules [29][30][31]. Numerous types of substituted benzoxazines have been synthesized via either cation-or radical-initiated tandem cyclizations [32][33][34][35]. Among them, selenylated benzoxazines have been successfully synthesized from the selenocyclization of selenium electrophiles, such as PhSeCl, PhSe + CF 3 COO − [36], and N-(PhSe)succinimide with o-vinylanilides [37] (Figure 2a). However, this approach is limited by the availability of selenium electrophiles (RSeX) and difficult to apply to R-modified selenocyclizations. Fe 3+ -promoted selenocyclization with readily available diorganyl diselenides could be a good alternative approach [38][39][40][41]. In our experiments (Figure 2b), excess FeCl 3 (2 equiv) was required to afford the desired products in 3−4 h due to the chelation of in situ-generated PhSe − with Fe 3+ . Since the generation of one PhSe + is accompanied by the formation of one PhSe − and the consumption of one molecule of Fe 3+ , the efficiency of this method is low in terms of atomic economy. It is worth noting that Zhang and coworkers reported that the combination of a catalytic amount of FeCl 3 /benzoyl peroxide (BPO) and 2 equiv of I 2 with diselenides afforded the desired products via both cationand radical-initiated pathways [42] (Figure 2c). According to the mechanism proposed by Zhang et al., BPO facilitated the generation of RSeI and RSe • , while FeCl 3 catalyzed the electrophilic cyclization of the neighboring aryl ring. More recently, Zhang et al. reported that I • generated from the redox reaction of FeCl 3 and KI induced the RSe • -initiated reaction pathway and the oxidation of the radical intermediates by Fe 3+ , and aerial oxygen yielded the desired products in 24 h under refluxing conditions [43] (Figure 2d). Inspired by these previous research, we envisioned that only catalytic amount of I 2 is actually needed to convert diselenide to RSeI, an ideal selenium electrophile [23,44], if the resulting I − could be recycled to I 2 by a second oxidant. Many high-valent metals are capable of oxidizing I − , but their low-valent counterparts are prone to be oxidized by O 2 . Therefore, the insertion of a multivalent metal redox cycle into the I 2 /I − redox cycle and O 2 may construct a doubleredox catalytic system for selenylated benzoxazines, featuring greener reaction conditions and high atomic economy. Herein, we report an efficient aerial O 2 -driven selenocyclization approach mediated by coupled Fe 3+ /Fe 2+ and I 2 /I − redox cycles ( Figure 2e). Mechanistic investigation confirmed the roles of each component in this novel double-redox catalytic system and revealed the reasons why Fe 3+ exhibited the best catalytic reactivity. due to the chelation of in situ-generated PhSe − with Fe 3+ . Since the generation of one PhSe + is accompanied by the formation of one PhSe − and the consumption of one molecule of Fe 3+ , the efficiency of this method is low in terms of atomic economy. It is worth noting that Zhang and coworkers reported that the combination of a catalytic amount of FeCl3/benzoyl peroxide (BPO) and 2 equiv of I2 with diselenides afforded the desired products via both cation-and radical-initiated pathways [42] (Figure 2c). According to the mechanism proposed by Zhang et al., BPO facilitated the generation of RSeI and RSe • , while FeCl3 catalyzed the electrophilic cyclization of the neighboring aryl ring. More recently, Zhang et al. reported that I • generated from the redox reaction of FeCl3 and KI induced the RSe • -initiated reaction pathway and the oxidation of the radical intermediates by Fe 3+ , and aerial oxygen yielded the desired products in 24 h under refluxing conditions [43] (Figure 2d). Inspired by these previous research, we envisioned that only catalytic amount of I2 is actually needed to convert diselenide to RSeI, an ideal selenium electrophile [23,44], if the resulting I − could be recycled to I2 by a second oxidant. Many high-valent metals are capable of oxidizing I − , but their low-valent counterparts are prone to be oxidized by O2. Therefore, the insertion of a multivalent metal redox cycle into the I2/I − redox cycle and O2 may construct a double-redox catalytic system for selenylated benzoxazines, featuring greener reaction conditions and high atomic economy. Herein, we report an efficient aerial O2-driven selenocyclization approach mediated by coupled Fe 3+ /Fe 2+ and I2/I − redox cycles ( Figure 2e). Mechanistic investigation confirmed the roles of each component in this novel double-redox catalytic system and revealed the reasons why Fe 3+ exhibited the best catalytic reactivity.

Results and Discussion
In the preliminary experiments, the catalytic reactivities of a series of multivalent transition metal salts, including Cu(acac) 2 , Co(acac) 3 , VO(acac) 2 , Ce(NH 4 ) 2 (NO 3 ) 6 , Ni(acac) 2 , MoO 2 (acac) 2 , PMA, Fe(acac) 3 , Fe(OTf) 3 , and FeCl 3 , were tested. The reaction of 1.0 equiv of o-vinylanilide 1a, 0.6 equiv of diselenide 2a, and 0.1 equiv of metal salt and I 2 in CH 3 CN was refluxed under air atmosphere. The data summarized in Table 1 (Entry 1−10) indicate that all of these multivalent transition metals showed some catalytic effect. A major problem with these metal catalysts is that these reactions had difficulty in terms of completion. Among these metals, Fe 3+ salts exhibited remarkably superior reactivity. In the presence of 10 mol% FeCl 3 , the selenocyclization took only 30 min to afford the desired benzoxazine product 3a with 93% yield. In addition, the experimental results (Table 1, Entry [11][12][13][14] showed that the FeCl 3 -catalyzed reaction exhibited notable solvent effects. Acetonitrile was proved as the most suitable solvent. In addition, we further reduced the amount of FeCl 3 and I 2 and found that the reaction could not go to completion when the amount of either FeCl 3 or I 2 was lowered to 5%.

Results and Discussion
In the preliminary experiments, the catalytic reactivities of a series of multivalent transition metal salts, including Cu(acac)2, Co(acac)3, VO(acac)2, Ce(NH4)2(NO3)6, Ni(acac)2, MoO2(acac)2, PMA, Fe(acac)3, Fe(OTf)3, and FeCl3, were tested. The reaction of 1.0 equiv of o-vinylanilide 1a, 0.6 equiv of diselenide 2a, and 0.1 equiv of metal salt and I2 in CH3CN was refluxed under air atmosphere. The data summarized in Table 1 (Entry 1−10) indicate that all of these multivalent transition metals showed some catalytic effect. A major problem with these metal catalysts is that these reactions had difficulty in terms of completion. Among these metals, Fe 3+ salts exhibited remarkably superior reactivity.
In the presence of 10 mol% FeCl3, the selenocyclization took only 30 min to afford the desired benzoxazine product 3a with 93% yield. In addition, the experimental results ( Table 1,  showed that the FeCl3-catalyzed reaction exhibited notable solvent effects. Acetonitrile was proved as the most suitable solvent. In addition, we further reduced the amount of FeCl3 and I2 and found that the reaction could not go to completion when the amount of either FeCl3 or I2 was lowered to 5%. With the optimized reaction conditions, we further explored the scope of this doubleredox catalytic system in the synthesis of selenylated benzoxazines. The data in Table 2 showed that the new method tolerated a variety of substituents on both o-vinylanilide and diorganyl diselenide substrates at different positions. Generally, both strong electronwithdrawing -NO 2 on the benzoyl moiety and more sterically hindered phenyl on the vinyl moiety resulted in lowered reaction rate. As shown in Table 2, selenylated benzoxazines 3a-3t were obtained in 80-93% yields in 0.5-2 h.
To elucidate the proposed roles of each key component in the catalytic system, a series of control experiments were performed. As shown in Scheme 1a,b, the absence of either catalytic amount of FeCl 3 or I 2 resulted in the formation of only a trace amount of 3a.
When the reaction was conducted in argon atmosphere, it simply afforded 3a in 38% yield (Scheme 1c). The above experimental results clearly indicate that FeCl 3 , I 2 , and aerial O 2 are essential to the catalytic system, but replacement of aerial O 2 with pure O 2 did not lead to improvement in the catalytic efficacy, suggesting that the proposed O 2 oxidation of Fe 2+ was not the rate-limiting step (Scheme 1d). Finally, the ineffectiveness of TEMPO on the proceeding of this reaction implies that radical species (RSe • ) should not be involved in reaction pathway [45,46] (Scheme 1e).  With the optimized reaction conditions, we further explored the scope of this double-redox catalytic system in the synthesis of selenylated benzoxazines. The data in Table 2 showed that the new method tolerated a variety of substituents on both o-vinylanilide and diorganyl diselenide substrates at different positions. Generally, both strong electron-withdrawing -NO2 on the benzoyl moiety and more sterically hindered phenyl on the vinyl moiety resulted in lowered reaction rate. As shown in Table 2, selenylated benzoxazines 3a-3t were obtained in 80-93% yields in 0.5-2 h. Table 2. Aerial O2-driven double-redox synthesis of selenylated benzoxazines (3a-3t).
To elucidate the proposed roles of each key component in the catalytic system, a series of control experiments were performed. As shown in Scheme 1a,b, the absence of either catalytic amount of FeCl3 or I2 resulted in the formation of only a trace amount of 3a. When the reaction was conducted in argon atmosphere, it simply afforded 3a in 38% yield (Scheme 1c). The above experimental results clearly indicate that FeCl3, I2, and aer-With the optimized reaction conditions, we further explored the scope of this double-redox catalytic system in the synthesis of selenylated benzoxazines. The data in Table 2 showed that the new method tolerated a variety of substituents on both o-vinylanilide and diorganyl diselenide substrates at different positions. Generally, both strong electron-withdrawing -NO2 on the benzoyl moiety and more sterically hindered phenyl on the vinyl moiety resulted in lowered reaction rate. As shown in Table 2, selenylated benzoxazines 3a-3t were obtained in 80-93% yields in 0.5-2 h. Table 2. Aerial O2-driven double-redox synthesis of selenylated benzoxazines (3a-3t).
To elucidate the proposed roles of each key component in the catalytic system, a series of control experiments were performed. As shown in Scheme 1a,b, the absence of either catalytic amount of FeCl3 or I2 resulted in the formation of only a trace amount of 3a. When the reaction was conducted in argon atmosphere, it simply afforded 3a in 38% yield (Scheme 1c). The above experimental results clearly indicate that FeCl3, I2, and aer- ial O2 are essential to the catalytic system, but replacement of aerial O2 with pure O2 did not lead to improvement in the catalytic efficacy, suggesting that the proposed O2 oxidation of Fe 2+ was not the rate-limiting step (Scheme 1d). Finally, the ineffectiveness of TEMPO on the proceeding of this reaction implies that radical species (RSe • ) should not be involved in reaction pathway [45,46] (Scheme 1e).

Scheme 1. Control experiments.
To prove the existence and coupling of the proposed Fe 3+ /Fe 2+ and I2/I − redox cycles, we further performed a series of control redox reactions without o-vinylanilide. As shown in Scheme 2a, a qualitative chromogenic assay ( Figure S1a and S1b) showed that the treatment of Fe 3+ with I − at ambient temperature caused an abrupt formation of a To prove the existence and coupling of the proposed Fe 3+ /Fe 2+ and I 2 /I − redox cycles, we further performed a series of control redox reactions without o-vinylanilide. As shown in Scheme 2a, a qualitative chromogenic assay ( Figure S1a,b) showed that the treatment of Fe 3+ with I − at ambient temperature caused an abrupt formation of a large amount of Fe 2+ and I 2 . In another experiment, the exposure of FeCl 2 in CH 3 CN to air at 80 • C for 10 min led to complete oxidation of Fe 2+ to Fe 3+ (Scheme 2b) as indicated by a chromogenic assay ( Figure S1c). Surprisingly, we found that almost no Fe 3+ could be detected after the aqueous solution of FeCl 2 was exposed to air for 30 min at 80 • C. This result indicates that the efficacy of the aerial oxidation of Fe 2+ is solvent dependent, which may possibly be one important reason why CH 3 CN is most suitable for this reaction. Finally, we tested the reaction of I 2 with PhSeSePh (Scheme 2c). As shown in Figure S2a, the reaction proceeded very slowly at ambient temperature. As indicated by the decoloration of I 2 , refluxing at 80 • C significantly promoted the formation of PhSeI. However, it still took 3 h for the reaction to reach equilibrium ( Figure S2b). When a catalytic amount of Fe 3+ (10 mol%) was added, the decoloration time was significantly shortened to 30 min ( Figure S2c). This result could be explained by the specific chelation of Fe 3+ with diselenide and induced polarization of Se-Se bonds, which have been reported in numerous previous reports [38][39][40][41]47]. The revealed dual roles of Fe 3+ well explained its superior catalytic activity compared with other multivalent metals in this selenocyclization reaction. On the basis of the above control experiments and tested redox reactions, a plausible mechanism was proposed for the current catalytic selenocyclization system ( Figure  3). At the beginning of the reaction, Fe 3+ catalyzed the formation of PhSeI (2a), the reactive selenium electrophile, which quickly reacted with o-vinylanilide (1a) and formed the seleniranium intermediate (INT1). The intramolecular nucleophilic cyclization (INT2) followed by the deprotonation by the superoxide radical anion (O2 •− ) afforded the desired benzoxazine (3a). Meanwhile, the released I − was instantly oxidized back to I2 by Fe 3+ to provide a continuous resource of PhSeI. The consumed Fe 3+ was also quickly recycled by the aerial oxidation of Fe 2+ . Therefore, the selenocyclization was pushed forward by a green oxidant, aerial oxygen, and only catalytic amounts of FeCl3 and I2 were needed. Since k2 > k1 > k3 in the redox cycles, the majority of ion metal existed as Fe 3+ , and iodine existed as I2 during the reaction process, which was verified by a chromogenic assay ( Figure S3).

General Methods
The solvents and chemical reagents used in the current research work were purchased from commercial suppliers. All of the reactions were monitored by TLC plates coated with 0.25 mm silica gel 60 F254 and visualized by 254 nm UV. The silica gel used in column chromatography (particle size 32-63 μm) was purchased from Qingdao Haiyang On the basis of the above control experiments and tested redox reactions, a plausible mechanism was proposed for the current catalytic selenocyclization system (Figure 3). At the beginning of the reaction, Fe 3+ catalyzed the formation of PhSeI (2a), the reactive selenium electrophile, which quickly reacted with o-vinylanilide (1a) and formed the seleniranium intermediate (INT1). The intramolecular nucleophilic cyclization (INT2) followed by the deprotonation by the superoxide radical anion (O 2 •− ) afforded the desired benzoxazine (3a). Meanwhile, the released I − was instantly oxidized back to I 2 by Fe 3+ to provide a continuous resource of PhSeI. The consumed Fe 3+ was also quickly recycled by the aerial oxidation of Fe 2+ . Therefore, the selenocyclization was pushed forward by a green oxidant, aerial oxygen, and only catalytic amounts of FeCl 3 and I 2 were needed. Since k 2 > k 1 > k 3 in the redox cycles, the majority of ion metal existed as Fe 3+ , and iodine existed as I 2 during the reaction process, which was verified by a chromogenic assay ( Figure S3). On the basis of the above control experiments and tested redox reactions, a plausible mechanism was proposed for the current catalytic selenocyclization system ( Figure  3). At the beginning of the reaction, Fe 3+ catalyzed the formation of PhSeI (2a), the reactive selenium electrophile, which quickly reacted with o-vinylanilide (1a) and formed the seleniranium intermediate (INT1). The intramolecular nucleophilic cyclization (INT2) followed by the deprotonation by the superoxide radical anion (O2 •− ) afforded the desired benzoxazine (3a). Meanwhile, the released I − was instantly oxidized back to I2 by Fe 3+ to provide a continuous resource of PhSeI. The consumed Fe 3+ was also quickly recycled by the aerial oxidation of Fe 2+ . Therefore, the selenocyclization was pushed forward by a green oxidant, aerial oxygen, and only catalytic amounts of FeCl3 and I2 were needed. Since k2 > k1 > k3 in the redox cycles, the majority of ion metal existed as Fe 3+ , and iodine existed as I2 during the reaction process, which was verified by a chromogenic assay ( Figure S3).

General Methods
The solvents and chemical reagents used in the current research work were purchased from commercial suppliers. All of the reactions were monitored by TLC plates

General Methods
The solvents and chemical reagents used in the current research work were purchased from commercial suppliers. All of the reactions were monitored by TLC plates coated with 0.25 mm silica gel 60 F 254 and visualized by 254 nm UV. The silica gel used in column chromatography (particle size 32-63 µm) was purchased from Qingdao Haiyang Chemicals, China. 1 H, 13 C, and 19 F NMR spectra were recorded on an AV-400 instrument (Bruker BioSpin, Faellanden, Switzerland) with chemical shifts referenced to DMSO-d 6 or CDCl 3 and reported in parts per million. Infrared spectra were obtained with a Vertex-70 instrument (Bruker Optics, Billerica, MA, US). HRMS spectra were acquired with a micrOTOF-Q II instrument (Bruker Daltonics, Billerica, MA, US)and reported as m/z. Melting points were measured on anX-4 melting point apparatus and were uncorrected (Tech Instrument, Beijing, China).The characterization data of new o-vinylanilides [37,[48][49][50][51] including 1e, 1f, 1g, and 1h and known selenylated benzoxazines [36,52]

Conclusions
In summary, a novel double-redox catalytic system was rationally constructed to provide efficient access to a variety of selenylated benzoxazines. The combination of only catalytic amounts of FeCl 3 and I 2 and the use of aerial oxygen as the end oxidant make this approach greener and more atomically efficient than conventional methods based on selenium electrophiles and FeCl 3 . This new method is widely applicable to a great diversity of o-vinylanilide and diorganyl diselenide substrates. Mechanistic investigation confirmed that the coupling of I 2 /I − and Fe 3+ /Fe 2+ catalytic redox cycles enabled aerial O 2 to act as the driving force to promote the selenocyclization reaction, which proceeds via a PhSeI-based pathway.