Efficient Synthesis of P-Chirogenic Compounds Enabled by Chiral Selenide-Catalyzed Enantioselective Electrophilic Aromatic Halogenation
Ruizhi Guo, Ziqi Liu, Xiaodan Zhao
Abstract
Open AccessCCS ChemistryCOMMUNICATION1 Nov 2021Efficient Synthesis of P-Chirogenic Compounds Enabled by Chiral Selenide-Catalyzed Enantioselective Electrophilic Aromatic Halogenation Ruizhi Guo, Ziqi Liu and Xiaodan Zhao Ruizhi Guo MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, Institute of Organic Chemistry, School of Chemistry, Sun Yat-Sen University, Guangzhou 510275 , Ziqi Liu MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, Institute of Organic Chemistry, School of Chemistry, Sun Yat-Sen University, Guangzhou 510275 and Xiaodan Zhao *Corresponding author: E-mail Address: [email protected] MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, Institute of Organic Chemistry, School of Chemistry, Sun Yat-Sen University, Guangzhou 510275 https://doi.org/10.31635/ccschem.020.202000530 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail P-chirogenic compounds have been applied in different fields, especially in asymmetric catalysis as ligands and organocatalysts. However, broader applicability has been severely restricted by the lack of efficient synthetic methods. Consequently, developing efficient methods to access these compounds is of high synthetic value. Herein, we report a convenient, efficient, and unprecedented pathway to construct valuable P-chirogenic compounds via chiral selenide-catalyzed enantioselective electrophilic aromatic halogenation. Using a new chiral bifunctional selenide as the catalyst, a variety of bis(2-hydroxyaryl) aryl phosphine oxides were efficiently converted to the corresponding chlorinated and brominated P-chirogenic compounds with good to excellent enantioselectivities. By slightly adjusting the catalyst and solvent, this method is also able to prepare chiral alkyl diaryl phosphine oxides and diaryl phosphinates. Furthermore, control experiments revealed the decomposition pathways of catalysts and the possible reasons why chiral selenide catalyst was more effective than chiral sulfide catalyst. The effect of hydrogen bonding was studied, and the reason why the chlorination took place on the various aromatic rings was elucidated when the substrates were switched from triaryl phosphine oxides to alkyl diaryl phosphine oxides and diaryl phosphinates. Download figure Download PowerPoint Introduction Phosphorus compounds bearing a P-stereogenic center are of great interest in the fields of agrochemicals and materials science.1,2 They have also been applied in asymmetric catalysis as ligands and organocatalysts because of the prominent chiral induction stemming from the chirality at the phosphorus atom closer to the catalytic center.3–6 Despite the importance of these compounds, a lack of efficient synthetic methods has largely restricted the breadth of applications.7–11 As a conventional synthetic route, P-chirogenic compounds were accessed on the basis of resolution of diastereomeric mixtures and chiral-auxiliary-based approaches. To avoid the consumption of stoichiometric amounts of chiral reagents, asymmetric catalytic methods such as metal-catalyzed asymmetric couplings12–20 and desymmetrization of prochiral phosphorus molecules have been developed.21–36 Particularly, desymmetrization methods have been a more focus lately. Typical examples include transition metal-catalyzed intra- and intermolecular C–H functionalization,22–28 Mo- and Ru-catalyzed ring-closing metathesis,29,30 Rh-catalyzed cycloaddition and hydroarylation,31,32 Au-catalyzed intramolecular hydroetherification,33N-heterocyclic carbene (NHC)-catalyzed esterification,34 Tm-catalyzed intermolecular sulfenylation,35 and cinchona alkaloid-catalyzed allylic alkylation reaction.36 However, these methods often suffer from substrate scope limitations. Some types of P-chiral compounds with practical potential such as triaryl phosphine oxides could not be achieved.34 Thus, developing new methods for efficient construction of challenging P-chirogenic molecules in a catalytic manner is highly desirable. Catalytic enantioselective electrophilic aromatic halogenation has emerged as an intriguing strategy for the synthesis of chiral molecules.37–51 By this kind of reaction, some valuable chiral halogenated molecules have been obtained. For instance, peptide-,37–43 chiral phosphoric acid-,44,45 or quinidine derivative-catalyzed46–48 enantioselective synthesis of axially chiral biaryls and benzamides via electrophilic aromatic bromination or iodination have been reported, and peptide-, enzyme- or amino-urea-catalyzed enantioselective synthesis of C-stereogenic methylenediphenols and methylenedianilines via electrophilic aromatic bromination or chlorination have been described (Figure 1a).49–51 Although this electrophilic reaction provides a convenient pathway for the synthesis of chiral aryl-substituted compounds, it has not been extensively explored due to the challenging issues of reactivity and control of enantioselectivity, and has never been applied to the construction of P-chirogenic compounds. Figure 1 | Construction of enantiopure molecules via catalytic enantioselective electrophilic reactions. (a) Known synthesis of chiral molecules via catalytic enantioselective electrophilic aromatic halogenation. (b) Chiral Lewis basic chalcogenide catalysis with electrophilic reagents. (c) This work: enantioselective synthesis of P-chirogenic compounds via chiral selenide-catalyzed electrophilic aromatic halogenation. Download figure Download PowerPoint Our group is interested in chiral Lewis basic chalcogenide catalysis.52–76 We noticed that this catalysis has been primarily applied to asymmetric electrophilic functionalization of alkenes such as enantioselective sulfenylation,52–63 bromination,65–67 trifluoromethylthiolation,68–74 and chlorination.75 Intermediates, chalcogenide-captured ion I and a chiral iranium ion II, were considered to be involved in the process (Figure 1b). In these electrophilic transformations, all obtained enantiopure products were generated on the basis of the construction of C-stereogenic centers via an iranium ion. With continued interest, we questioned whether chiral compounds could be synthesized by chalcogenide catalysis to avoid the construction of an iranium ion, namely, whether chiral compounds could be formed by enantioselective electrophilic aromatic substitution reaction in which the formed chalcogenide-captured ion I was attacked by the aromatic ring of prochiral aryl-substituted molecules. If this proposed strategy is feasible, P-chirogenic compounds might be formed from prochiral organophosphorus molecules by similar chiral chalcogenide-catalyzed electrophilic aromatic halogenation. Herein, we report our discovery of chiral bifunctional selenide catalyst-enabled enantioselective desymmetrizing electrophilic aromatic halogenation. By this protocol, a variety of P-chirogenic triaryl phosphine oxides, alkyl diaryl phosphine oxides, and diaryl phosphinates were efficiently synthesized (Figure 1c). Results and Discussion Condition optimization We initiated enantioselective synthesis of P-chirogenic compounds via electrophilic aromatic halogenation with readily available bis(2-hydroxy-5-methoxyphenyl) phenyl phosphine oxide ( 1a) as the model substrate (Table 1 and see the preliminary results in Supporting Information Scheme S1). The hydroxy group on the substrate was expected not only to increase the electron density of the phenyl group, but also to serve as a H-bonding donor or acceptor to interact with the catalyst. Owing to the convertibility of the chloro group and the success of chiral sulfide-catalyzed enantioselective chlorination of alkenes,75 chiral bifunctional sulfide C1 was tested for the chlorination of 1a in the presence of N-chlorosuccimide (NCS). No reaction took place at −78 °C although electrophilic chlorination of phenols could proceed to form the chlorinated products at room temperature with NCS (entry 1).77–80 Similarly, chlorinating reagent 1,3-dichloro-5,5-dimethylhydantoin (DCDMH) did not give the desired product either (entry 2). To trigger chlorination, more reactive N-chlorosaccharin (NCSA) was utilized. Pleasingly, the chlorination occurred to give the desired product, albeit in trace amounts, and dichlorinated byproduct 3 was not observed (entry 3). However, in the reaction, a chlorinated sulfoxide 4 derived from the sulfide catalyst was detected. It was rationalized that the sulfide-captured chlorenium ion may be unable to react with 1a, and then decomposed to sulfoxide 4. Table 1 | Screening of Reaction Conditions Entry Catalyst Cl+ Source (equiv) Solvent Yield (%) ee of 2a (%) 2a 3 1 C1 NCS (1.1) CH2Cl2 — 0 — 2 C1 DCDMH (1.1) CH2Cl2 — 0 — 3 C1 NCSA (1.1) CH2Cl2 6 0 9 4 C2 NCS (1.1) CH2Cl2 — — — 5 C2 DCDMH (1.1) CH2Cl2 Trace — — 6 C2 NCSA (1.1) CH2Cl2 61 8 87 7 C2 NCSI (1.1) CH2Cl2 34 10 62 8 C2 tBuOCl (1.1) CH2Cl2 59 4 80 9 C3 NCSA (1.1) CH2Cl2 60 6 92 10 C4 NCSA (1.1) CH2Cl2 50 13 54 11 C5 NCSA (1.1) CH2Cl2 68 7 93 12 C6 NCSA (1.1) CH2Cl2 72 6 95 13 C6 NCSA (1.1) Toluene 2 11 — 14 C6 NCSA (1.1) THF 80 4 98 15 C6 NCSA (1.3) THF 83 6 99 16 C6 NCSA (1.5) THF 80 16 >99 17a C6 NCSA (1.3) THF 89 10 >99 18a ent -C6 NCSA (1.3) THF 84 12 −98 19a — NCSA (1.3) THF 24 Trace — Conditions: 1a (0.025 mmol), Cl+ source (1.1–1.5 equiv), catalyst (10 mol %), solvent (1.0 mL), −78 °C, 12 h. Yield was determined by 31P NMR with triphenylphosphine oxide as the internal standard. The ee value was determined by HPLC analysis on a chiral stationary phase. aFor 1.5 days. To tune the reactivity of the chlorenium ion, selenide catalyst C2 was examined for the reaction. Almost no reaction occurred with NCS or DCDMH (entries 4 and 5). The decomposition of C2 to 3-triflylaminoindene ( 5) was observed in the reactions. To our surprise, when NCSA was used, the desired product 2a associated with dichlorinated byproduct 3 was formed by enantioselective chlorination in 61% yield with 87% ee (entry 6). Other highly reactive chlorinating reagents such as tert-butyl hypochlorite (tBuOCl) and N-chlorodibenzenesulfonimide (NCSI) led to lower yields and ees (entries 7 and 8). These results drove us to focus on the modification of selenide catalyst based on C2. When catalyst C3 bearing an ortho-substituted methyl group on the phenyl ring was used, enantioselectivity increased (entry 9). However, more sterically hindered, ortho-disubstituted catalyst C4 gave the product only in moderate yield and ee (entry 10). After a series of selenide catalysts were screened (see Supporting Information Scheme S2), it was found that catalyst C5 bearing a methyl group at the ortho position and a methoxy group at the para position of the phenyl ring largely facilitated the reaction (entry 11). By slight adjustment, a new selenide C6 bearing an ortho-methoxy group and a para-isopropoxy group as the catalyst gave the optimized result (entry 12). Other reaction parameters were evaluated as well. The chlorination was dramatically affected by solvent polarity. The chlorination almost ceased in toluene and the best result was obtained with tetrahydrofuran (THF) as the solvent (entries 13 and 14). By adjusting the amount of NCSA to 1.3 equiv, the yield of 2b increased to 83% with 99% ee (entry 15). However, further increases resulted in lower yields due to the generation of the more dichlorinated byproduct 3 (entry 16). To our delight, the best result was obtained by C6 as catalyst in the presence of 1.3 equiv of NCSA in THF for 1.5 days (entry 17). The enantiomer of catalyst C6 was also able to catalyze the chlorination to give the enantiomer of 2a in high yield with −98% ee (entry 18). It is worth noting that the uncatalyzed reaction was sluggish and only 24% yield of the chlorinated product was observed after 1.5 days (entry 19). Synthesis of P-chirogenic triaryl phosphine oxides With optimal conditions in hand, we turned our attention to investigate the substrate scope of enantioselective electrophilic aromatic halogenation of bis(2-hydroxyaryl) aryl phosphine oxides (Table 2). The phenol moieties bearing different electron-donating groups at the five-position were well tolerated regardless of size, and the desired products were formed in high yields with excellent enantioselectivities ( 2a–2e, >99% ee each). This reaction was sensitive to the electron density of phenol moieties. When electron-withdrawing groups such as the fluoro group were introduced into the same position, the chlorination became sluggish and high levels of starting materials remained even if the amount of chlorine increased or highly reactive chlorinating reagents were used (see Supporting Information Scheme S3). Moreover, when bis(2-hydroxyphenyl)phenyl phosphine oxide without a substituent at the five-position of the phenol moiety was used as substrate, the chlorination gave the desired monochlorinated product only in low yields with excellent enantioselectivity because of the formation of multichlorinated byproducts. Interestingly, sterically congested substrates bearing a multisubstituted phenol moiety underwent chlorination efficiently to give products with excellent enantioselectivities ( 2f, >99% ee; 2g, 99% ee). It is worth mentioning that the position isomer of 1a (3-MeO group on the phenol moieties) was compatible with this protocol. In this case, the chlorination inversely took place in the five-position to generate the product with high enantioselectivity as well ( 2ha, 51% yield, 97% ee). Furthermore, variation of the aryl group was investigated. A variety of electron-rich and electron-deficient aromatic rings were suitable for the reaction ( 2i–2p, 51–99% yields, 83–99% ee). Increasing steric hindrance led to a dramatic yield decrease due to the formation of more dichlorination byproducts ( 2l, 51% yield). Of note, in some cases, 1.5 equiv of NCSA was used to ensure full conversion of starting materials ( 2e, 2o, and 2p) or the product underwent dimethylation of the –OH groups for the improved isolation and detection of the enantiomeric excess ( 2c and 2h). This robust catalysis was also suitable for enantioselective aromatic bromination of triaryl phosphine oxides. By replacing the halogen source and solvent, the brominated product was obtained at the same level of success ( 6, 96% ee). Table 2 | Scope of Enantioselective Electrophilic Aromatic Halogenation of Bis(2-Hydroxyaryl) Aryl Phosphine Oxides Conditions: 1 (0.10 mmol), NCSA (1.3 equiv) or N-bromosaccharin (NBSA) (1.3 equiv), THF (4.0 mL), −78 °C, 1.5 days. Yield is isolated yield. Ratio of ee was determined by HPLC analysis on a chiral stationary phase. aRefer to the isolated yield and ee value of the product from the methylation of the corresponding 2c or 2h. bNCSA (0.15 mmol) was used. cCH2Cl2 (4.0 mL) + Cl2CHCHCl2 (4.0 mL) as the solvent. Synthesis of P-chirogenic alkyl diaryl phosphine oxides and diaryl phosphinates To demonstrate the generality of our developed protocol, the chlorination of methyl diaryl phosphine oxide 7, which is an important precursor of 1,2-bis(o-anisylphenylphosphine)ethane (DiPAMP) analogues was investigated.81 However, the yield and enantioselectivity of the product were low when the reaction was performed under the standard conditions with C6 as catalyst (Figure 2a). We considered that this unsatisfied result might stem from the improper steric hindrance of catalyst C6 (see Supporting Information Scheme S4). Consequently, various chiral selenide catalysts were screened. It was found that a more sterically hindered catalyst C4 was most efficient for the conversion of 7. In contrast, when the diastereomer of catalyst C4 (dia- C4) was used as catalyst, the product with the same absolute configuration at the phosphorus center was still generated with 34% (see Supporting Information Scheme S5). Using catalyst C4, different solvents were screened. Finally, the chlorination proceeded smoothly to produce product 8 in moderate yield with good enantioselectivity in THF and toluene mixed solvents (see Figure 2a and Supporting Information Scheme S6). More importantly, this modified condition was effective for asymmetric aromatic chlorination of diaryl phosphinates (Figure 2b). Different P-chirogenic diaryl phosphinates were obtained in moderate to good yields with good to excellent enantioselectivities ( 10a–10f, 63–98% ee). It is noteworthy that the chlorination occurred regioselectively on the phenol moiety although electron-rich para-methoxyphenyl group was on the substrate ( 10d). The absolute configuration of chiral phosphinates 10 was determined to be S based on the X-ray crystallographic analysis of 10e (see Supporting Information Figure S6 and relevant Figures S3– S5). Interestingly, the chlorination took place on the opposite phenol moiety compared with the reaction of aryl-substituted phosphine oxides 1. Figure 2 | Synthesis of P-chirogenic methyl diaryl phosphine oxide and diaryl phosphinates. (a) Enantioselective aromatic chlorination of methyl diaryl phosphine oxide. (b) Enantioselective aromatic chlorination of diaryl phosphinates. Conditions: 9 (0.10 mmol), NCSA (1.3 equiv), CH2Cl2 (4.0 mL), −78 °C, 1.5 days. Yield is isolated yield. Ratio of ee was determined by high-performance liquid chromatography (HPLC) analysis on a chiral stationary phase. aNCSA (0.15 mmol) was used. bTHF/toluene = 4∶1 (4.0 mL) were used. cRefer to the isolated yield and ee value of the product from the methylation of the OH groups of the corresponding 10f. Download figure Download PowerPoint Further transformations of products To test the practicability of this catalytic system, the subgram-scale reaction was tested. For example, chlorinated product 2ca, which was generated from 1c in the process of enantioselective chlorination and then dimethylation could be scaled up to 1 mmol (0.34 g) without yield or enantioselectivity erosion (Figure 3a). In addition, the obtained product 2ca could be further converted into different P-chirogenic phosphorus compounds via Pd-catalyzed cross-coupling reaction (Figure 3b). Compared with aryl bromides and iodides, the cross-coupling reaction of aryl chlorides was generally more challenging due to the lower inherent reactivity. To overcome this issue, accelerating the oxidative addition process with electron-rich ligands is the key.82 We found that the bulky electron-rich ligands, S-Phos and X-Phos, were highly efficient for the Suzuki–Miyaura coupling and Sonogashira coupling of phosphine oxide 2ca, respectively. When 2ca was treated with boronic acids and phenylacetylene in the presence of suitable palladium precatalyst and ligands, the desired cross-coupling products were generated in high yields without erosion of enentioselectivity ( 11– 14). Furthermore, the chiral phosphine oxide 13 could be efficiently reduced to give product 15 bearing reverse chirality at the phosphorus atom.83 Figure 3 | Transformations of products. (a) Subgram-scale reaction. (b) Conversion of 2ca via Pd-catalyzed cross-coupling reaction. Download figure Download PowerPoint Mechanism insights In a previous work,75 it was found that sulfides as catalysts were effective for the electrophilic chlorination of alkenes, but selenide catalysts were not. In this reaction, the effects of sulfide and selenide catalysts were inverse. Why did this happen? Based on the experimental results in Table 1 and our previous studies,75 we rationalized that inverse effects might be attributed to the difference in reactivity between sulfonium chloride I and selenonium chloride II (Figure 4a). Under these conditions, the sulfide catalyst was unable to activate the Cl+ source, and the generated sulfonium chloride I was relatively inert. So, the sulfide catalyst was incapable of aromatic substitution. As a result, the formed sulfonium chloride I was quenched by trace amounts of water from solvent and reagents to give sulfoxide. Then the sulfoxide underwent further chlorination to form 4 (see Supporting Information Figure S7), but unable to generate 5 via elimination of the sulfur moiety. In contrast, the selenide catalyst was superior in activating the Cl+ source, and selenonium chloride II was reactive enough for the aromatic substitution. Due to its high reactivity, it was possible to slowly decompose to selenoxide. Moreover, the selenoxide was unstable and preferred to undergo elimination of the selenium moiety to form 5 rather than further chlorination (see Supporting Information Figure S2). Besides, the fact that sulfoxide 5 was isolated as a single diastereoisomer reflected that sulfonium chloride I and selenonium chloride II might be formed with excellent diastereoselectivity. The excellent diastereoselectivity meant that sulfur- and selenium-based catalytic centers had excellent chirality, which could benefit the enantioselective formation of products in the subsequent electrophilic reaction. Figure 4 | Mechanistic insights. (a) The decomposition pathways of selenide catalyst and sulfide catalyst. (b) Kinetic resolution of racemic 2a. (c) The effects of substrate and catalyst. (d) The favored and disfavored pathways. Download figure Download PowerPoint During the reactions, besides monochlorinated products, racemic dichlorinated byproducts generated from the further chlorination of products were always detected. This side reaction might go through a kinetic resolution process. To understand this point, the kinetic resolution of chlorination of racemic 2a was studied. When the chlorination of racemic 2a was carried out with catalyst C6 and 0.65 equiv of NCSA, 2a was recovered in 52% yield with 75% ee (Figure 4b). This result indicated that further chlorination was a kinetic resolution process. However, as shown in Table 1, 2a was formed with >99% ee associated with dichlorinated byproduct 3 in 10% yield under the optimal conditions. Based on the result that the kinetic resolution process only gave 2a with relatively low enantioselectivity, we considered that the enantioselectivity of reaction largely originated from the desymmetrizing aromatic halogenation rather than the kinetic resolution process (see Supporting Information Figure S1). Next, we tried to understand how the products were formed enantioselectively in the aromatic chlorination. According to previous work,75 H-bonding effects between the substrate and catalyst generally play an important role for stereocontrol. To elucidate the influence of this effect in the reaction, control experiments were conducted. When both of the –OH groups on substrate 1a were converted to methoxy groups, it was surprising to find that the reaction was shut down under standard conditions (Figure 4c). When substrate 1a with free –OH groups underwent chlorination in the presence of N -Me-C6 derived from catalyst C6 by methyl protection on the nitrogen, the chlorination reaction took place, but gave the racemic product (Figure 4c). These results indicate that the H-bonding donors from substrate and catalyst are crucial for the reaction: the –OH group on the substrate is essential for the occurrence of electrophilic aromatic chlorination at low temperature, and the trifluoromethylsulfonamido (–NHTf) group on the catalyst benefits the yield and determines the enantioselective formation of the product. On the basis of the above results and the literature,71,75 the possible mechanistic pathway for formation of 2a is proposed, as shown in Figure 4d. First, a selenide-captured chlorenium ion Int-I is formed by reaction of Cl+ source with selenide catalyst. Then, it interacts with one of the phenol moieties of 1a by an anion bridge. By means of this interaction, the phenol group is activated to be reactive enough for the following chlorination. During the interaction, intermediate Int-II is preferred as the phenyl group of 1a faces to the aryl group of catalyst. If the phenyl group of 1a faces the opposite direction, the formed intermediate Int-II′ is not preferred due to the steric hindrance. In contrast, when diaryl phosphinates, such as 9a, are utilized as substrates, Int-IV is preferred owing to less steric hindrance compared with Int-IV′ with the methoxy group facing the aryl group of the catalyst. Finally, the subsequent aromatic chlorination gives the desired products via Int-III. Conclusions We developed a new and convenient route to access P-chirogenic triaryl phosphine oxides via chiral bifunctional selenide-catalyzed enantioselective aromatic halogenation with bis(2-hydroxyaryl)aryl phosphine oxides. Most of the desired products were obtained in high yields with excellent enantioselectivities. This method was also suitable for the synthesis of chiral methyl diaryl phosphine oxide and diaryl phosphinates. Moreover, the products could be converted to more complex P-chiral compounds via Pd-catalyzed cross-coupling reaction. Results from the control experiments revealed the decomposition pathway of the catalyst and indicated that H-bonding was essential for the occurrence of reaction and control of enantioselectivity. The stereocontrol of reactions with different types of substrates was elucidated. This work is greatly complementary to the field of enantioselective electrophilic aromatic substitution as well as the synthesis of P-chirogenic compounds and will guide chiral chalcogenide catalysis toward a new direction in asymmetric synthesis. The application of this electrophilic strategy in other reactions and the practical applications of these P-chiral compounds in asymmetric reactions are ongoing in our Supporting Information Supporting Information is of is no of interest to The the of and the for the and the and of for 1. Phosphorus in and Phosphorus in Chiral Phosphorus for Enantioselective 4. and Phosphorus on a for and of P-Chirogenic Phosphorus as Chiral to 7. of Phosphorus Synthesis of Chiral Enantioselective Synthesis of in Synthesis of Phosphorus at Phosphorus on the Synthetic for the of Enantioselective Synthesis of a P-Chirogenic Catalytic Synthesis of via a