Catalytic Enantioselective Construction of Chiral Benzo-Fused <i>N</i> -Heterocycles through Friedel–Crafts-Type Electrophilic Chlorination

Abstract

Open AccessCCS ChemistryCOMMUNICATION1 May 2022Catalytic Enantioselective Construction of Chiral Benzo-Fused N-Heterocycles through Friedel–Crafts-Type Electrophilic Chlorination Jie Luo, Yuanyuan Zhang, Fuming Zhong and Xiaodan Zhao Jie Luo Institute of Organic Chemistry and MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, School of Chemistry, Sun Yat-sen University, Guangzhou 510275 , Yuanyuan Zhang Institute of Organic Chemistry and MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, School of Chemistry, Sun Yat-sen University, Guangzhou 510275 , Fuming Zhong Institute of Organic Chemistry and MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, School of Chemistry, Sun Yat-sen University, Guangzhou 510275 and Xiaodan Zhao *Corresponding author: E-mail Address: [email protected] Institute of Organic Chemistry and MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, School of Chemistry, Sun Yat-sen University, Guangzhou 510275 https://doi.org/10.31635/ccschem.021.202100777 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Chiral benzo-fused N-heterocycles are frequently found in natural and synthetic products. However, their synthesis usually suffers from different limitations such as difficulty in accessing appropriate starting materials and unsatisfactory stereoselectivities. In this work, an unprecedented chiral sulfide-catalyzed enantioselective Friedel–Crafts-type electrophilic chlorination is shown to construct various 3,4-functionalized tetrahydroquinolines with excellent enantio- and diastereoselectivities from readily available aniline derivatives. Interestingly, employing N-allyl 1-naphthanilides as substrates, divergent reactions via chlorocarbocyclization and dearomatization occurred to afford two chiral polycyclic benzo-fused N-heterocycles. The system that we developed extends the scope of asymmetric chlorination to general substrates without the need of a N–H group, and significantly promotes the synthesis of enantioenriched benzo-fused N-heterocycles. Download figure Download PowerPoint Introduction Chiral benzo-fused nitrogen heterocycles are an important class of compounds with unique properties that are widely used in the pharmaceutical and agrochemical industries.1–4 In past decades, many methods have been developed for their synthesis to meet various requirements in different fields.5–9 Despite tremendous progress, the current methods are still subject to various limitations. For example, chiral tetrahydroquinolines10–13 belong to the family of these heterocycles, and many efforts have been made on their synthesis. However, the methods developed usually suffer from the functionalization of only specific functional groups at the 2-, 3-, or 4-position of the tetrahydroquinolines formed, the scarcity of materials, and the unsatisfactory enantioselectivity of the reactions (Figure 1a).14–27 On the other hand, it remains difficult to produce benzo-fused N-heterocycles with more rings, such as chiral polycyclic rings using the current systems.28,29 Therefore, it is desirable to develop efficient methods to synthesize diverse chiral benzo-fused N-heterocycles from readily available substrates. Figure 1 | Catalytic enantioselective construction of chiral N-heterocycles. (a) Known methods toward asymmetric synthesis of tetrahydroquinolines. (b) Known asymmetric chlorination of allyl amides. (c) No report about asymmetric Friedel–Crafts-type electrophilic chlorination. (d) Construction of chiral tetrahydroquinolines and N-polycyclic heterocycles through asymmetric Friedel–Crafts-type electrophilic chlorination. Cat., catalyst. *, symbol of chiral center. Download figure Download PowerPoint The catalytic enantioselective electrophilic functionalization of alkenes has become an ideal strategy for the synthesis of various chiral molecules.30–35 In this transformation, alkene substrates react with electrophilic iodine, bromine, chlorine, or sulfur reagents to form a reactive iranium ion intermediate, which is then attacked by tethered or external nucleophiles to yield multifunctionalized products. To realize the reactions with different types of alkenes, various metal- and organocatalyzed methods have been developed.30–40 Particularly, due to the good convertibility of the chlorine atom and its prevalence in natural products and bioactive compounds, as well as the more reactive nature of the chloriranium ion that promotes intriguing transformations compared with other iranium ions such as brom-, iod-, and thiiranium ions, catalytic asymmetric electrophilic chlorination of alkenes has received considerable attention from chemists in the past decade.41–47 For instance, Borhan et al.41–45 reported hydroquinidine 1,4-phthalazinediyl diether [(DHQD)2PHAL]-catalyzed asymmetric chlorolactonization of olefinic carboxylic acids, and developed intra- and intermolecular chlorination of allyl amides with the same catalyst (Figure 1b). In addition, other groups have developed catalytic asymmetric chlorination of allyl alcohols,33,36 homoallylic alcohols,46 2-vinylphenylcarbamates,47 and so on.48,49 However, as a facile route to synthesize a variety of chiral tetrahydroquinolines, catalytic asymmetric Friedel–Crafts-type electrophilic chlorination of easily prepared N-allyl anilines with electrophilic chlorine reagent has not been demonstrated (Figure 1c). It is possible that this cyclization is very challenging due to the following problems: (1) Known methods for asymmetric chlorination rely on the use of secondary amides with a necessary N–H group that can act as a hydrogen-bonding donor to interact with the chiral catalyst to help control the chiral environment of the reaction.42–45,49–54 When protected N-allyl secondary anilines, namely tertiary N-allyl anilines, are used as substrates, not only the steric hindrance on the nitrogen increases, but also the hydrogen bonding from the nitrogen-containing group disappears, which theoretically results in hard control of the enantioselectivity of the reaction. (2) Using tertiary N-allyl aniline derivatives as substrates, the electron-rich aromatic ring of the substrates easily undergoes electrophilic substitution,55–57 which can retard the functionalization of the double bond to form the desired products. Owing to these issues, the synthesis of racemic tetrahydroquinolines from protected N-allyl anilines was achieved only under specific electrophilic reaction conditions.58–62 Herein, we report our discovery that by using the easily removable α-hydroxyacetyl group as the protecting group (PG), N-allyl anilides can efficiently yield tetrahydroquinolines via chiral sulfide-catalyzed Friedel–Crafts-type electrophilic chlorination. Interestingly, by employing similar N-allyl 1-naphthanilides as substrates, divergent reactions occurred to afford two different chiral polycyclic N-heterocycles (Figure 1d). Results and Discussion Optimization of reaction conditions Based on previous experience with asymmetric reactions catalyzed by chiral chalcogenides,49,63–71 we set out to test asymmetric chlorocarbocyclization of N-cinnamyl aniline compound 1 using chiral chalcogenide catalysts in the presence of acid (Table 1). Generally, allyl anilide 1 was easily prepared from the corresponding anilines in two steps. Using benzoyl as the PG, the desired product was formed in 44% yield with only 12% ee using 1,3-dichloro-5,5-dimethylhydantoin (DCDMH) as the chlorinating source and chiral sulfide C1 as the catalyst (entry 1). We note that no arene-chlorinated product was observed under these conditions. Switching the chlorinating reagent to N-chlorosuccinimide (NCS) improved the ee to 30%, but led to poor reactivity. Encouragingly, when the chlorinating reagent 1,3-dichlorohydatoin (DCH) was employed in the reaction, both reactivity and enantioselectivity were improved (entry 3, 60% yield, 30% ee). Based on these preliminary results, we turned our attention to the effect of different PGs on the substrate. Acyl, tosyl, and acetoxyacetyl (PG1) groups led to lower reactivity and worse enantioselectivity of the transformation than the benzoyl group (entries 4–6). Surprisingly, the 2-hydroxyacetyl group (PG2) with a free OH improved the enantioselectivity to 54% ee. The result led us to try more PGs with tethered H-bonding donor groups, for example, PG3 and PG4 groups with a tertiary alcohol and sulfonamide, respectively. However, neither of them led to higher enantioselectivity (entries 8 and 9). Then, different chalcogenide catalysts were tested. Catalyst C2 with both methyl and methoxy groups on the aryl ring improved the enantioselectivity of the reaction to 68% ee (entry 10). In contrast, the selenide catalyst C3 was not catalytically active in this reaction (entry 11). Satisfactorily, using sulfide catalyst C4 with a bulky isopropyl group, the product was formed in 90% ee, albeit in a lower yield (entry 12). After screening the reaction solvents, we found that a mixed solvent of chloroform and pentane evidently enhanced the reactivity and the selectivity of the reaction, affording the desired product in 84% yield and 96% ee (entry 15).72,73 In addition, reducing the catalyst loading led to the decrease of the yield of the product (entry 16). It should be noted that acids played an important role in the reactions (entries 17–22). For example, when BF3·OEt2 was used, both the yield and the enantioselectivity were significantly decreased. Further, almost no reaction was observed when MsOH, TsOH, Sc(OTf)3, or no acid was added. Table 1 | Screening of Reaction Conditions Entry Cat. Cl+ reagent Additive Pg Yield (%)a ee (%)b 1 C1 DCDMH TMSOTf Bz 44 12 2 C1 NCS TMSOTf Bz 28 30 3 C1 DCH TMSOTf Bz 60 30 4 C1 DCH TMSOTf Ac 22 12 5 C1 DCH TMSOTf Ts <5 — 6 C1 DCH TMSOTf PG1 <5 — 7 C1 DCH TMSOTf PG2 25 54 8 C1 DCH TMSOTf PG3 9 32 9 C1 DCH TMSOTf PG4 23 3 10 C2 DCH TMSOTf PG2 56 68 11 C3 DCH TMSOTf PG2 <5 — 12 C4 DCH TMSOTf PG2 26 90 13c C4 DCH TMSOTf PG2 90 92 14d C4 DCH TMSOTf PG2 8 66 15e C4 DCH TMSOTf PG2 84 96 16e,f C4 DCH TMSOTf PG2 46 95 17e C4 DCH Tf2NH PG2 80 94 18e C4 DCH BF3•Et2O PG2 28 74 19e C4 DCH MsOH PG2 <5 — 20e C4 DCH TsOH PG2 <5 — 21e C4 DCH Sc(OTf)3 PG2 <5 — 22e C4 DCH — PG2 <5 — Reaction conditions: 1a (0.05 mmol), Cl+ reagent (1.5 equiv), catalyst (15 mol %), additive (1.0 equiv), CH2Cl2 (2.0 mL), −78 °C, 18 h. Without note, dr < 20:1. aNMR yield using quinoline as the internal standard. bDetermined by HPLC analysis. cCHCl3 (2.0 mL) as solvent at −60 °C. dPentane (2.0 mL) as solvent. eCHCl3 (2.0 mL) + pentane (2.0 mL) as solvents. f5 mol % catalyst was used. Construction of chiral tetrahydroquinolines The above results indicated a facile route to synthesize chiral functionalized tetrahydroquinolines with high selectivities from 2-hydroxyacetyl group-protected anilides. Importantly, the PG could be easily removed. Using glycol as solvent, product 2a could be easily converted into free tetrahydroquinoline with 93% yield at 60 °C in 24 h without the erosion of enantioselectivity (Table 2, 2a′). In addition, under optimal conditions, anilides with PG3 and PG4 could also be transformed to the corresponding products with high yields and ees ( 2b, 86% ee; 2c, 91% ee). The influence of substituents on the aromatic ring of the cinnamyl group was then investigated in consideration of steric hindrance and electron effect. All of the reactions proceeded well to give the desired products in high yields with excellent enantioselectivities despite the presence of methyl, chloro, or bromo group in the substrates ( 2d– 2h, 94–98% ees). Even if 1f with an ortho-methyl group on the phenyl ring was used as the substrate, the reaction still produced the product efficiently ( 2f, 98% ee). We note that, using substrate with a heterocyclic ring such as the thiophene ring on the double bond, no product was detected under the standard conditions. Similarly, the effect of substituents on the phenyl ring of the aniline unit was also studied. When the substituents such as methyl, halogen, and phenyl groups were placed in the para-position, the reactivity of the reactions did not change. All of the products were produced in excellent enantioselectivies ( 2i– 2l, 94–98% ees). However, when the substituents were placed at the ortho-position, the reactions became sluggish, possibly because of steric hindrance in the cyclization step. Nevertheless, the anilides with an alkyl group or halogen atom such as bromine at the meta-position underwent cyclization to form the desired products in excellent regio-, enantio-, and diastereoselectivities ( 2m– 2p, 94–96% ees). We note that cyclopropyl group was not affected under these conditions ( 2o, 90% yield). Furthermore, when two halogen atoms such as fluorine and chlorine were installed on the phenyl ring, the corresponding product 2q was still generated in good yield and high regioselectivity with high ee. Besides, 2-naphthanilide as the substrate afforded the desired product 2r in 97% yield with 94% ee. Unexpectedly, when the anilides with gem-dimethyl-substituted alkene were employed as substrates, enantioselective hydroxychlorination rather than chlorocarbocyclization occurred to yield chiral chlorinated γ-amino alcohols with high ees (Table 2, 2s and 2t). This result shows that it is possible for 2-hydroxyacetyl-protected anilides to take other pathways to yield different types of products. Table 2 | Enantioselective Construction of Chiral Tetrahydroquinolines with Different N-Allyl Anilides Conditions: 1 (0.1 mmol), DCH (1.5 equiv), TMSOTf (1.0 equiv), catalyst C4 (15 mol %), CHCl3 (4.0 mL) + pentane (4.0 mL), −78 °C, 18 h. Without note, dr < 20:1. aConcentrated HCl (0.5 mL), glycol (1 mL), 60 °C, 24 h. bH2O (2 equiv) was added. Divergent reactions of allyl 1-naphthanilides Interestingly, by employing 1-nathphanilide 3a as substrate under optimal conditions, two products of the reaction were detected (Figure 2a). Of these two products, the structure of oxazolidinone-based polycyclic N-heterocycle 4a, which was formed by the dearomatization of the naphthalene ring, was characterized by nuclear magnetic resonance (NMR), and further confirmed by X-ray crystallographic analysis to have four continuous chiral centers (41% yield, 92% ee, and excellent diastereoselectivity). Notably, such nitrogen heterocyclic moieties are easily found in natural alkaloids.74–78 To date, although dearomatization of phenols and naphthols has been extensively investigated and successfully applied to the synthesis of various chiral ketone derivatives,79,80 it is still difficult to achieve asymmetric dearomatization of aromatic amines.81 Another product 5a was characterized as benzo[h]tetrahydroquinolines (46% yield, 93% ee, and excellent diastereoselectivity). Due to the hindered rotation of the amide group in the cyclization product, the signals of 5a in NMR appear as a mixture of rotamers at ambient temperature. However, the structure of 5a was confirmed by variable temperature NMR experiments, and its absolute configuration was further determined by X-ray crystallographic analysis of its derivative 5a′ through easy deprotection of the PG. Figure 2 | Divergent reactions of 1-naphthanilides. (a) Two products generated starting from 3a. (b) Kinetic resolution of 6a. (c) Studies on formation of diastereomer of 4b. Download figure Download PowerPoint The formation of two different products motivated us to find the reason for the special phenomenon. Interestingly, when the amount of chlorinating reagent was reduced to 0.6 equiv, products 4a and 5a were generated in the same yields along with 27% of 3a remaining. The result supported that there was no kinetic difference between the formation of the two products. Besides, using PhSPh as the catalyst, products 4a and 5a were generated as well, albeit in relatively lower yields, which indicated that the divergent reactions were possibly a substrate-controlled process instead of a catalyst-controlled process. Additionally, high-performance liquid chromatography (HPLC) analysis of substrate 3a on a chiral stationary phase indicated that 3a was unstably atropisomeric since two atropisomers could not be completely separated by HPLC at ambient temperature. However, considering the reaction was carried out at −78 °C, the atropisomerism of the substrate might become stronger, and render divergent reactions of two atropisomers to form two different products.82–84 X-ray study of 3a also manifested that the amide plane was orthogonal to the naphthalene ring in the solid state, which supported the possibility of atropisomerism (Figure 2a). Moreover, a substrate with a relatively higher rotation barrier was synthesized. In this case, kinetic resolution was observed using substrate 6a with 2-methyl substituent (Figure 2b). The dearomatized product 7a bearing two continuous chiral quaternary carbon centers was formed in good enantioeselectivity, and chiral substrate (S)- 6a was recovered with 93% ee in the presence of 0.6 equiv of DCH in a mixed solvent of CHClCCl2 and PhF. In this case, the chlorocarbocyclization took place at the ipso-C-1 position first, possibly because of the electron-donating substituent effect in electrophilic dearomatization of arylamides (also see Figure 3).81 It was worth mentioning that increasing the amount of chlorinating reagent promoted the further conversion of (S)- 6a to dia- 7a. Figure 3 | Divergent reaction of electron-rich naphthanilide 6b. Download figure Download PowerPoint To gain insight into the formation of two products in this reaction, further efforts were made to improve the efficiency of the nucleophilic attack of the hydroxy group. Based on the Thorpe–Ingold effect, two methyl groups were introduced into the carbon chain connected to the hydroxy group of naphthanilide 3b. Gratifyingly, when the reaction was carried out under similar conditions, the diastereomer dia- 4b was observed along with the formation of 4b, albeit in <15% yield (Figure 2c, aromatized cyclization product was observed as well). X-ray crystallographic studies revealed that dia- 4b with syn-configuration of the phenyl ring and 1,2-dihydronaphthalene ring was much more sterically congested than product 4b. Thus, in the case of 3a, such congestion might lead to the formation of the less sterically hindered product 5a via β-H elimination from one atropisomer rather than the formation of the analogue of the dearomatized product dia -4b (also see Proposed mechanism section). Based on the above results, divergent reactions of 1-naphthanilides were evaluated (Table 3). When different aryl substituents were placed on the double bond, both the corresponding dearomatized products and benzo[h]tetrahydroquinoline derivatives were obtained with excellent stereoselectivities (entries 1–8). In this transformation, steric hindrance on the double bond did not affect the reaction (entry 1). To achieve full conversion of different substrates, the concentration of the solution and the reaction time were slightly adjusted. For example, when an electron-withdrawing group such as F–, Cl–, Br–, or CF3O– was placed on the phenyl ring of the styryl moiety, longer reaction time was required for complete conversion (entries 3 and 6–8). In general, the yields of products 5 were higher than the yields of products 4. In some cases, the yields of product 5 even exceeded 50%. This result indicated that the reactions were prone to give product 5. In light of our proposal, the β-H elimination might slightly surpass the nucleophilic attack of the hydroxy group when they compete with each other. Furthermore, halo substituents such as Br– at the 4- or 5-position on the naphthalene ring did not have a large impact on the formation of products with high enantioselectivities (entries 9 and 10). In one attempt, when the aryl group on the double bond of 3 was replaced by alkyl substituents, the catalytic method was not effective for these substrates. In addition, when electron-donating group such as cyclopropyl group was placed at the 4-position of the naphthalene ring, two dearomatized polycyclic 7b and dia- 7b could be obtained with high enantioselectivities (Figure 3). Table 3 | Divergent Reactions of 1-Naphthanilides via Chlorocarbocyclization and Dearomatization Conditions: 1 (0.1 mmol), DCH (1.5 equiv), TMSOTf (1.0 equiv), catalyst C4 (15 mol %), CHCl3 (4 mL) + pentane (4 mL), −78 °C, 12 h. Without note, dr < 20:1. aCHCl3 (2 mL) + pentane (2 mL). b48 h. cdr = 10∶1. ddr = 15∶1. To further enhance the efficiency to synthesize benzo[h]tetrahydroquinolines, we wondered whether sole benzo[h]tetrahydroquinoline products could be formed under different conditions. Interestingly, when Lewis acids, for example, Sc(OTf)3, Zn(OTf)2, Ln(OTf)3, and Cu(OTf)2, were added into the reactions, product 5a was generated as the major product. Using 0.5 equiv of Cu(OTf)2, 5a was produced in 90% yield without the erosion of enantioselectivity (Figure 4). Since Cu(OTf)2 was unable to directly promote the conversion of 4a to 5a under the reaction conditions (see the Supporting Information), it is possible that the copper cation can coordinate the oxygen atom of the hydroxy group and the carbonyl group of 3a to retard the nucleophilic attack of the hydroxy group. Figure 4 | Generation of sole product from 1-naphthanilide by the addition of additives. Download figure Download PowerPoint With these results in hand, we examined the functional groups for the synthesis of benzo[h]tetrahydroquinolines, and all of the products were obtained in excellent enantioselectivities (Table 4). There was no big difference between the use of substrates with ortho- or meta-substituents on the phenyl ring. However, 1 equiv of acid TMOTf was required for the substrates bearing weak electron-withdrawing groups (Cl, OCF3, and Br) to ensure the high yields of the products. These results showed an efficient avenue to synthesis of multisubstituted chiral benzo[h]tetrahydroquinolines. Table 4 | Exploration of Functional Groups on the Synthesis of Benzo[h]Tetrahydroquinolines Conditions: 3 (0.1 mmol), DCH (1.5 equiv), TMSOTf (0.5 equiv), C4 (15 mol %), CHCl3 (4 mL) + pentane (4 mL), −78 °C, 12 h. aTMSOTf (1.0 equiv) was used. Figure 5 | Proposed mechanism for divergent reactions of 1-naphthanilides. Download figure Download PowerPoint Proposed mechanism Based on these experimental results and previous studies,67 plausible reaction pathways are depicted for the asymmetric synthesis of chiral benzo-fused N-heterocycles using the system developed. The reactions starting from 1-naphthanilides were selected as the example to explain the formation of two exquisite products in one step (Figure 5). First, at low temperatures, substrate 3a, comprising two C–N atropisomers (S)- 1 and (R)- 1, reacts with the cationic complex I,71,85 which is formed from chlorinating reagent DCH in the presence of catalyst C4 and TMSOTf, to construct two diastereomers int-I and int-II with a chiral chlorinium ion and an anion bridge derived from the corresponding acid. The anion bridge between the hydroxy group of the PG and the catalyst is proposed to help control the chiral environment of the whole transformation, explaining the crucial role of the PG in the reaction. Then, the aromatic ring attacks the chlorinium ion intramolecularly to give int-III and int-IV in the process of axial chirality transfer to point chirality along with the release of catalyst C4. Finally, nucleophilic attack of the hydroxy group on int-III toward the cation leads to the dearomatization of the naphthalene ring to give product 4a and TfOH prior to β-H elimination. In contrast, int-IV gives product 5a and TfOH via β-H elimination due to the congestion between the phenyl ring and the 1,2-dihydro-naphthalene ring (about the congestion, see Figure 2c and the We have developed an efficient method to synthesize chiral benzo-fused N-heterocycles from readily available through chiral sulfide-catalyzed enantioselective electrophilic In this work, an easily removable 2-hydroxyacetyl group was introduced into the substrate to interact with the catalyst to help control the Interestingly, employing N-allyl 1-naphthanilides as substrates, not only chiral benzo[h]tetrahydroquinolines but also oxazolidinone-based polycyclic were obtained in one step under similar conditions. With the addition of Cu(OTf)2 sole chiral benzo[h]tetrahydroquinolines could be obtained with excellent This a route for the synthesis of chiral and is to the of electrophilic of alkenes and asymmetric of chiral chalcogenide for other challenging reactions is in our Supporting Supporting is available and experimental HPLC and NMR of of There is no of to The the of and the for the and the and of for Chemistry in for and of the and of 4. in 5. of and Using in the Synthesis of of and in the Chemistry of The of on and and from Luo from an from the and from of Dearomatization Reaction of

Topics & Concepts