Synthesis of Ester-Containing Chroman-4-Ones via Cascade Radical Annulation of 2-(Allyloxy)Arylaldehydes with Oxalates under Metal Free Conditions

A convenient and practical method for the synthesis of bioactive ester-containing chroman-4-ones through the cascade radical cyclization of 2-(allyloxy)arylaldehydes and oxalates is described. The preliminary studies suggest that an alkoxycarbonyl radical might be involved in the current transformation, which was generated via the decarboxylation of oxalates in the presence of (NH4)2S2O8.


Introduction
The construction of ester-containing compounds is among the most valuable transformations in modern organic synthesis, given the abundance of these motifs in many functional materials, natural products, and pharmaceuticals [1][2][3][4], to name a few ( Figure 1). They are also important building blocks which can be converted into versatile functional groups such as carboxyl, hydroxymethyl, aldehyde, amide, etc. [5,6]. Consequently, developing novel and practical methods for the synthesis of organic molecules containing such structural moiety has received extensive attention [7][8][9][10]. Traditionally, the synthesis of esters relies on the esterification of carboxylic acids, acyl chlorides or anhydrides with various alcohols. These approaches need the pre-installation of a carboxyl group in the substrate. Doubtless, the direct introduction of an ester group to organic compounds represents a more effective strategy. In this context, transition-metal-catalyzed alkoxycarbonylation using carbon monoxide and alcohol as ester sources might be an attractive and alternative protocol [11][12][13]. However, this method often suffers from the use of toxic CO and highpressure equipment. Thus, the development of alkoxycarbonylation reactions with readily available and easily handled esterification reagents other than carbon monoxide is still in high demand. sources [24]. Recently, Wu and Chen reported an Ir(ppy)3 catalyzed alkoxycarbonylation/cyclization reaction of N-acryloyl benzamides with alkyloxalyl chlorides under visible light irradiation [25]. Despite these achievements, developing more alkoxycarbonyl radical triggered cascade cyclization reactions is still highly desirable. On the other hand, chroman-4-ones are among the ubiquitous structural motifs that occur in natural products, pharmaceuticals and biologically active compounds. They exhibit a variety of physiological and biological activities, such as antibacterial, antioxidant, anti-HIV, and SIRT2 inhibiting properties [26][27][28][29][30]. For example, 8-bromo-6-chloro-2-pentylchroman-4-one was selected as a potent inhibitor of SIRT2 with an IC50 of 1.5 μM in a reported work in the literature [30]. In the past few years, various radical cascade annulation reactions of 2-(allyloxy)arylaldehydes triggered by diverse radicals, such as alkyl, acyl, phosphoryl, trifluoromethyl and sulfonyl, for the synthesis of valuable chroman-4one derivatives have been well established (Scheme 1a) [31][32][33][34][35][36][37][38][39][40][41][42]. Nevertheless, to the best of our knowledge, an alkoxycarbonyl radical triggered cascade radical cyclization to synthesize ester-functionalized chroman-4-ones has never been reported. Herein, we disclose In recent years, cascade radical annulation has emerged as an efficient strategy for the synthesis of functionalized heterocycles. A variety of novel and practical radicals triggered cascade annulation reactions have been widely reported [14][15][16][17][18][19]. Among these, ester radical induced annulation reactions have also gained great attention [20,21]. For instance, in 2013, Du and Li reported an iron-catalyzed oxidative radical cascade annulation of N-aryl acrylamides with carbazates toward alkoxycarbonylated oxindoles [22]. In 2014, Zhu et al., reported an iron-catalyzed radical alkoxycarbonylation of 2-isocyanobiphenyl with carbazates leading to phenanthridine-6-carboxylates [23]. In 2017, Sun and coworkers reported a visible light induced cascade radical cyclization reaction toward ester-functionalized pyrido [4,3,2-gh]phenanthridind derivatives using carbazates as the ester sources [24]. Recently, Wu and Chen reported an Ir(ppy) 3 catalyzed alkoxycarbonylation/cyclization reaction of N-acryloyl benzamides with alkyloxalyl chlorides under visible light irradiation [25]. Despite these achievements, developing more alkoxycarbonyl radical triggered cascade cyclization reactions is still highly desirable.
On the other hand, chroman-4-ones are among the ubiquitous structural motifs that occur in natural products, pharmaceuticals and biologically active compounds. They exhibit a variety of physiological and biological activities, such as antibacterial, antioxidant, anti-HIV, and SIRT2 inhibiting properties [26][27][28][29][30]. For example, 8-bromo-6-chloro-2pentylchroman-4-one was selected as a potent inhibitor of SIRT2 with an IC 50 of 1.5 µM in a reported work in the literature [30]. In the past few years, various radical cascade annulation reactions of 2-(allyloxy)arylaldehydes triggered by diverse radicals, such as alkyl, acyl, phosphoryl, trifluoromethyl and sulfonyl, for the synthesis of valuable chroman-4-one derivatives have been well established (Scheme 1a) [31][32][33][34][35][36][37][38][39][40][41][42]. Nevertheless, to the best of our knowledge, an alkoxycarbonyl radical triggered cascade radical cyclization to synthesize ester-functionalized chroman-4-ones has never been reported. Herein, we disclose a (NH 4 ) 2 S 2 O 8 -mediated protocol for the selective intramolecular decarboxylative radical cyclization of 2-(allyloxy)arylaldehydes with oxalates for the rapid building of a variety of ester-containing alkyl-substituted chroman-4-ones (Scheme 1b). It is worth mentioning that oxalates can be easily obtained via the condensation of readily available alcohols and oxalyl chloride, followed by in situ hydrolysis without tedious column chromatography, making the present method much more practical and attractive to give various ester-containing chroman-4-ones.

Results and Discussion
Initially, a model reaction of 2-(allyloxy)benzaldehyde (1a) and 2-methoxy-2-oxoacetic acid (2a) was carried out to investigate the reaction conditions (Table 1). When the reaction was performed in DMSO at 80 °C under N2 atmosphere for 24 h employing (NH4)2S2O8 (3 equiv.) as the oxidant, the desired product, methyl 2-(4-oxochroman-3yl)acetate (3aa), was obtained in 71% isolated yield (entry 1). Inspired by the above result, some other common solvents including DMF, CH3CN, DCE, THF and H2O were also screened. To our surprise, only DMSO was efficient for the present transformation and no desired 3aa was observed with other examined solvents (entries 2-6). Furthermore, several mixed solvents consisting of different volume ratios of DMSO and H2O were tested

Results and Discussion
Initially, a model reaction of 2-(allyloxy)benzaldehyde (1a) and 2-methoxy-2-oxoacetic acid (2a) was carried out to investigate the reaction conditions (Table 1). When the reaction was performed in DMSO at 80 • C under N 2 atmosphere for 24 h employing (NH 4 ) 2 S 2 O 8 (3 equiv.) as the oxidant, the desired product, methyl 2-(4-oxochroman-3-yl)acetate (3aa), was obtained in 71% isolated yield (entry 1). Inspired by the above result, some other common solvents including DMF, CH 3 CN, DCE, THF and H 2 O were also screened. To our surprise, only DMSO was efficient for the present transformation and no desired 3aa was observed with other examined solvents (entries 2-6). Furthermore, several mixed solvents consisting of different volume ratios of DMSO and H 2 O were tested (entries 7-10). The results revealed that a trace amount of H 2 O was more efficient for the current reaction (entry 7 vs. entry 1), and 3aa was isolated with 76% yield in DMSO/H 2 O (500:1). Next, various oxidants including K 2 S 2 O 8 , Na 2 S 2 O 8 , TBHP, Selectfluor and PhI(OAc) 2 were also investigated (entries [11][12][13][14][15]. In sharp contrast to (NH 4 ) 2 S 2 O 8 , no desired product 3aa was detected with the other examined oxidants. Moreover, the influence of reaction temperature to the reaction was then tested. Elevating the temperature from 80 • C to 90 • C, the yield of 3aa increased from 76% to 81% (entry 16 vs. entry 7); further raising the temperature to 100 • C gave a slightly reduced yield of 3aa (entry 17). When the reaction temperature was reduced from 90 • C to 70 • C, 3aa was obtained with an obviously decreasing yield (entry 18). In addition, reducing the amount of (NH 4 ) 2 S 2 O 8 or 2a to 2 equiv. (based on 1a), the desired 3aa was obtained in 57 and 64% yields, respectively (entries [19][20]. Moreover, in the absence of any oxidant, no reaction occurred, suggesting that oxidant was crucial for the current reaction (entry 21).  With the established optimal reaction conditions ( Table 1, entry 16), we first investigated the generality of this cascade annulation reaction by employing various substituted 2-(allyloxy)arylaldehydes with 2-methoxy-2-oxoacetic acid (2a). As depicted in Scheme 2, the substrates bearing either electron-donating substitutions (Me, OMe, t-Bu and OBn) or electron-withdrawing substitutions (F, Cl, Br and CO2Me) at the different positions of the With the established optimal reaction conditions ( Table 1, entry 16), we first investigated the generality of this cascade annulation reaction by employing various substituted 2-(allyloxy)arylaldehydes with 2-methoxy-2-oxoacetic acid (2a). As depicted in Scheme 2, the substrates bearing either electron-donating substitutions (Me, OMe, t-Bu and OBn) or electron-withdrawing substitutions (F, Cl, Br and CO 2 Me) at the different positions of the benzene ring all reacted smoothly to give the desired products in moderate to good yields (3aa-3na). Furthermore, the di-substituted and naphthalene derived substrates were also transformed into the desired products 3pa and 3qa in moderate yields. Moreover, substrate 1r, bearing a methyl substitution close to C=C bond and 2-allylbenzaldehyde 1s, also reacted well, giving the corresponding products 3ra and 3sa in 53 and 52% yields, respectively. However, N-allyl-N-(2-formylphenyl)acetamide 1t as a substrate failed to give any of the desired product under the current condition (3ta). Next, the reaction scope with respect to oxalates was evaluated. As shown in Scheme 3, various oxalates (2b-2j) with varied aliphatic chains containing diverse substitutions, including alkyl (2b-2f), phenyl (2g), fluorine atom (2h), alkoxyl (2i) and trifluoromethyl (2j), proceeded well to deliver the desired products 3ab-3aj in good yields. In addition, cyclic substituted substrates were also compatible with the reaction, giving 3ak and 3al in Next, the reaction scope with respect to oxalates was evaluated. As shown in Scheme 3, various oxalates (2b-2j) with varied aliphatic chains containing diverse substitutions, including alkyl (2b-2f), phenyl (2g), fluorine atom (2h), alkoxyl (2i) and trifluoromethyl (2j), proceeded well to deliver the desired products 3ab-3aj in good yields. In addition, cyclic substituted substrates were also compatible with the reaction, giving 3ak and 3al in 81 and 53% yields, respectively. Moreover, using 2-(tert-butoxy)-2-oxoacetic acid 2m as a substrate, a more stable tert-butyl radical was generated via a double decarboxylation process, which further reacted with 1a to access the corresponding product 3am in 78% yield. substrate, a more stable tert-butyl radical was generated via a double decarboxylation process, which further reacted with 1a to access the corresponding product 3am in 78% yield. To illustrate the synthetic application of this method, a gram scale experiment of 1a and 2a was performed under standard conditions and the desired product 3aa was obtained in 77% yield (Scheme 4a). Additionally, further valuable transformations of 3aa were carried out. First, 3aa could be smoothly hydrolyzed to carboxyl-containing compound 4a under acidic conditions with an excellent yield (Scheme 4b). Moreover, the treatment of 3aa with CeCl3 and NaBH4 gave the cis-lactone 5a in 67% yield as the exclusive product [43,44] (Scheme 4c). To illustrate the synthetic application of this method, a gram scale experiment of 1a and 2a was performed under standard conditions and the desired product 3aa was obtained in 77% yield (Scheme 4a). Additionally, further valuable transformations of 3aa were carried out. First, 3aa could be smoothly hydrolyzed to carboxyl-containing compound 4a under acidic conditions with an excellent yield (Scheme 4b). Moreover, the treatment of 3aa with CeCl 3 and NaBH 4 gave the cis-lactone 5a in 67% yield as the exclusive product [43,44]  To better understand this cascade cyclization process, some control experim were performed as shown in Scheme 5. When the model reaction was conducted un the standard reaction conditions with the addition of 2 equiv. of BHT or TEMPO as free radical inhibitor, no desired product 3aa was observed and TEMPO-adduct 4b detected by GC-MS analysis (Scheme 5a), indicating that a free-radical pathway migh involved in the present transformation. Moreover, a reaction of 1a and 2a was perform in the presence of 1,1-diphenylethene under standard conditions, the radical adduct resulting from the trapping of the initial methyl ester radical by 6a was observed, wh further confirmed that the current reaction was triggered by the alkoxycarbonyl rad generated through the decarboxylation of oxalates during the reaction (Scheme 5b).  To better understand this cascade cyclization process, some control experiments were performed as shown in Scheme 5. When the model reaction was conducted under the standard reaction conditions with the addition of 2 equiv. of BHT or TEMPO as the free radical inhibitor, no desired product 3aa was observed and TEMPO-adduct 4b was detected by GC-MS analysis (Scheme 5a), indicating that a free-radical pathway might be involved in the present transformation. Moreover, a reaction of 1a and 2a was performed in the presence of 1,1-diphenylethene under standard conditions, the radical adduct 6aa resulting from the trapping of the initial methyl ester radical by 6a was observed, which further confirmed that the current reaction was triggered by the alkoxycarbonyl radical generated through the decarboxylation of oxalates during the reaction (Scheme 5b). To better understand this cascade cyclization process, some control experiments were performed as shown in Scheme 5. When the model reaction was conducted under the standard reaction conditions with the addition of 2 equiv. of BHT or TEMPO as the free radical inhibitor, no desired product 3aa was observed and TEMPO-adduct 4b was detected by GC-MS analysis (Scheme 5a), indicating that a free-radical pathway might be involved in the present transformation. Moreover, a reaction of 1a and 2a was performed in the presence of 1,1-diphenylethene under standard conditions, the radical adduct 6aa resulting from the trapping of the initial methyl ester radical by 6a was observed, which further confirmed that the current reaction was triggered by the alkoxycarbonyl radical generated through the decarboxylation of oxalates during the reaction (Scheme 5b).

General Information
Unless otherwise specified, all chemicals were purchased from commercial suppliers and directly used as received without additional purification. Column chromatography was carried out with silica gel (200-300 mesh) to purify products, using proper solvents as the eluent system. NMR spectra were recorded at 400 MHz for 1 H NMR spectra and 100 MHz for 13 C NMR spectra by using a German Bruker Avance 400 spectrometer. Chemical shifts are quoted in parts per million referenced to the appropriate solvent peak ( 1 H NMR: CDCl3 7.26 ppm, 13 C NMR: CDCl3 77.0 ppm). The following abbreviations were used to describe peak splitting patterns when appropriate: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet. Mass spectra were performed on a spectrometer operating on ESI-TOF.

General Information
Unless otherwise specified, all chemicals were purchased from commercial suppliers and directly used as received without additional purification. Column chromatography was carried out with silica gel (200-300 mesh) to purify products, using proper solvents as the eluent system. NMR spectra were recorded at 400 MHz for 1 H NMR spectra and 100 MHz for 13 C NMR spectra by using a German Bruker Avance 400 spectrometer. Chemical shifts are quoted in parts per million referenced to the appropriate solvent peak ( 1 H NMR: CDCl 3 7.26 ppm, 13 C NMR: CDCl 3 77.0 ppm). The following abbreviations were used to describe peak splitting patterns when appropriate: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet. Mass spectra were performed on a spectrometer operating on ESI-TOF.

General Procedure for the Synthesis of Target Products 3
An oven-dried 10 mL reaction tube was charged with 2-(allyloxy)arylaldehyde 1 (0.3 mmol, 1 eq), oxalates 2 (0.9 mmol, 3 eq) and (NH 4 ) 2 S 2 O 8 (0.9 mmol, 3 eq) in a DMSO aqueous solution (1.8 mL, V DMSO /V H2O = 500/1) with a magnetic stirring bar. The mixture was then stirred at 90 • C under N 2 atmosphere conditions for about 24 h. The reaction was monitored by TLC. After completion, water (10 mL) was added and the mixture was extracted with EtOAc (10 mL × 3); the solvent was then removed under vacuum. The residue was purified by flash column chromatography using a mixture of petroleum ether and ethyl acetate as eluent to give the desired products 3.

Gram-Scale Synthesis of 3aa
An oven-dried 100 mL round-bottom flask was charged with 2-(allyloxy)benzaldehyde 1a (0.972 g, 6 mmol), 2-methoxy-2-oxoacetic acid 2a (1.872 g, 18 mmol) and (NH 4 ) 2 S 2 O 8 (4.104 g, 18 mmol) in a DMSO aqueous solution (36 mL, V DMSO /V H2O = 500/1) with a magnetic stirring bar. The mixture was then stirred at 90 • C under N 2 atmosphere conditions for about 24 h. The reaction was monitored by TLC. After completion, water (30 mL) was added and the mixture was extracted with EtOAc (40 mL × 3); the solvent was then removed under vacuum. The residue was purified by flash column chromatography using a mixture of petroleum ether and ethyl acetate as eluent to give 1.016 g of 3aa, yielding 77%.