Metalloenzymes can promote a wide range of oxidative transformations, such as hydroxylation, epoxidation, and cis-dihydroxylation. Inspired by the properties of metalloenzymes, researchers have established biomimetic synthetic metal (e.g., Mn, Fe) complexes coordinated with tetradentate nitrogen (N4) ligands as excellent catalysts in oxidation reactions in the past decades [1-4]. The landmark investigations of epoxidation of olefins by biomimetic N4 metal complexes were begun in 2001 by Jacobsen and co-workers [5]. They reported that an Fe(Ⅱ)(mep) complex (mep = N, N′-dimethyl-N, N′-bis (2-pyridinylmethyl)ethane-1, 2-diamine) could rapidly mediate the epoxidation of aliphatic alkenes with aqueous hydrogen peroxide as the oxidant; note that acetic acid served as an important additive in the iron catalytic system. Earlier in the same year, Que and co-workers [6] first demonstrated enantioselective cis-dihydroxylation of olefins catalyzed by the iron complex of the aminopyridine N4 ligand. In 2003, Mn-MCP-(OTf)₂ [MCP = N, N-dimethyl-N, N-bis(2-pyridylmethyl) cyclohexane-trans-1, 2-diamine] was used in the epoxidation of olefins with peracetic acid as an oxidant by Stack and co-workers; importantly, an ee of 10% was observed in the epoxidation of vinyl cyclohexane [7]. The development of novel elegant ligands is considered the key approach to obtaining highly efficient catalysts [8, 9]. In this context, Costas [10-14], Talsi and Bryliakov [15-20], Gao [21, 22], and our group [23-31] have developed many chiral N4 ligands, and their manganese and iron complexes exhibit good to excellent stereocontrol in the enantioselective epoxidation of various olefins (Scheme 1). Among them, the chiral N4 ligands derived from chiral 2, 2′-bipyrrolidine play an important role in iron- or manganese-catalyzed asymmetric epoxidation of olefins (Scheme 1)[11, 12, 14, 16-20]. However, these chiral 2, 2′-bipyrrolidine skeletons are not easily available and are costly. In addition, the epoxidation catalyzed by manganese or iron complexes coordinated by these documented ligands requires a large quantity of acids as a vital additive [16, 23]. It has been proposed that protonation of the hydroperoxide ligand by the coordinated carboxylic acid in metal-hydroperoxo intermediates facilitates O–O bond cleavage, generating high-valent metal-oxo species as reactive epoxidizing intermediates. Recently, Costas and co-workers reported that the quantity of carboxylic acid could be dramatically decreased to catalytic loading by including dimethylamino groups at the 4-position of the PDP ligand [Scheme 1, ^NMe2PDP, PDP = 2-({2-[1-(pyridin-2-ylmethyl)-pyrrolidin-2-yl]pyrrolidin-1-yl}methyl) pyridine] [12, 13]. In 2012, we reported a series of facile N4 ligands (Scheme 1, S-PMP, S-PEB) derived from L-proline, which exhibited enantioselectivity comparable to those of chiral 2, 2′-bipyrrolidine skeletons [28-32]. More importantly, L-proline is abundant in nature and easily transformed and diversified.

On the basis of our previous work, we hypothesized that modification of the substituents on the pyridyl groups and diamine backbone of the N4 ligand from L-proline would provide several new and practical N4 ligands. Herein, we report three structurally new aminopyridine ligands derived from L-proline and the catalytic performance of the corresponding manganese complexes in asymmetric epoxidation of a variety of olefins using aqueous hydrogen peroxide as an oxidant.

The starting materials were purchased from commercial suppliers and used without further purification. ¹H NMR and ¹³C NMR spectra were recorded on a Bruker Avance Ⅲ 400 MHz spectrometer using CDCl₃ as the solvent with tetramethylsilane as an internal reference. Gas chromatography–mass spectrometry (GC–MS) spectra were recorded on an Agilent Technologies 7890A GC system with an Agilent 5975 inert mass-selective detector (EI) and an HP-5MS column (0.25 mm × 30 m, film: 0.25 μm). High-performance liquid chromatography (HPLC) analysis was performed on a Waters-Breeze instrument (2487 Dual λ absorbance detector and 1525 binary HPLC pump). Chiralpak OD-H, AS-H, and IC columns were purchased from Daicel Chemical Industries, Ltd. GC analysis was performed on an Agilent 6820 GC instrument with a CP-Chirasil-Dex CB column. Column chromatography was generally performed on silica gel (300–400 mesh), and thin-layer chromatography inspections were performed on silica gel GF₂₅₄ plates.

L-proline-based diamines 1 and 2 were synthesized using a previously reported method [28, 29, 33].

For L1, (S)-N-methyl-1-(pyrrolidin-2-yl)methanamine (1) (2.1 mmol) and 30 mL of dichloromethane (DCM) were added to a 100-mL round-bottom flask. Then K₂CO₃ (6 eq.) was added. Next, 2-(chloromethyl)-N, N-dimethylpyridin-4-amine (3) (5.0 mmol) was dissolved in 20 mL of DCM by addition through a constant-pressure dropping funnel at 0 ℃. Then the mixture was stirred in an ice bath for 1 h and gradually warmed to room temperature overnight. The mixture was then filtered, and the filter was washed with DCM. The combined filtrates were evaporated under reduced pressure. To the resulting residue, a solution of NaOH (2 mol/L) was added to adjust the pH to greater than 10, and the mixture was extracted with 3 × 20 mL of DCM. The combined organic phases were washed successively with saturated aqueous solutions of NaCl and H₂O. The organic phase was dried over anhydrous Na₂SO₄, and the solvent was removed under reduced pressure. The residue was purified by a silica column to yield the ligand L1 (330 mg, 41%). ¹H NMR (400 MHz, CDCl₃) δ 8.10–7.94 (m, 2H), 6.60 (d, J = 8.0 Hz, 2H), 6.31 (dd, J = 8.0, 4.0 Hz, 2H), 4.17 (d, J = 12.0 Hz, 1H), 3.50 (dt, J = 28.0, 16.0 Hz, 3H), 2.89 (dd, J = 12.0, 4.0 Hz, 12H), 2.82 (t, J = 8.0 Hz, 2H), 2.58–2.47 (m, 1H), 2.39–2.25 (m, 2H), 2.21 (s, 3H), 2.02–1.87 (m, 1H), 1.7–1.51 (m, 3H). ¹³C NMR (101 MHz, CDCl₃) δ 158.2, 157.9, 155.2, 155.1, 147.6, 147.5, 105.8, 105.4, 105.3, 105.2, 64.1, 62.5, 62.2, 60.9, 54.7, 43.2, 39.2, 39.1, 29.8, 23.0. High-resolution MS (HRMS) [electrospray ionization (ESI) MS] calcd. for C₂₂H₃₅N₆ [M+H]⁺: 383.2918, found: 383.2924.

L2 was prepared using a method analogous to that used for L1 starting with (S)-N-benzyl-1-(pyrrolidin-2-yl) methanamine (2) (2.1 mmol) and 2-(chloromethyl)-N, N-dimethylpyridin-4-amine (6) (5.0 mmol) to provide a yellow oil (356 mg, 37%). ¹H NMR (400 MHz, CDCl₃) δ 7.62 (t, J = 4.0 Hz, 4H), 7.43–7.12 (m, 5H), 6.41 (d, J = 2.0 Hz, 1H), 6.34 (dd, J = 8.0, 4.0 Hz, 1H), 4.33 (d, J = 8.0 Hz, 1H), 4.24–4.01 (m, 2H), 3.66–3.50 (m, 3H), 2.99 (s, 12H), 2.81 (dd, J = 12.0, 4.0 Hz, 2H), 2.67 (dd, J = 12.0, 8.0 Hz, 1H), 2.37 (dd, J = 12.0, 8.0 Hz, 1H), 2.05–1.80 (m, 2H), 1.76–1.60 (m, 3H), 1.61–1.42 (m, 1H). ¹³C NMR (101 MHz, CDCl₃) δ 156.1, 155.3, 143.2, 132.1, 130.2, 128.8, 105.2, 104.7, 61.4, 59.1, 54.4, 53.5, 51.7, 51.3, 39.7, 29.1, 23.9. HRMS (ESI-MS) calcd. for C₂₈H₃₉N₆ [M+H]⁺: 459.3231, found: 459.3218.

L3 was prepared using a method analogous to that used for L1 starting with (S)-N-methyl-1-(pyrrolidin-2-yl) methanamine (1) (2.1 mmol) and 2-(chloromethyl)- 4-methoxy-3, 5-dimethylpyridine (4) (5.0 mmol) to yield the ligand L3 (398 mg, 46%). ¹H NMR (400 MHz, CDCl₃) δ 8.07 (s, 2H), 4.04 (d, J = 12.0 Hz, 1H), 3.65 (d, J = 8.0 Hz, 6H), 3.46 (dd, J = 44.0, 12.0 Hz, 3H), 2.76 (s, 1H), 2.59 (s, 1H), 2.36 (dd, J = 12.0, 4.0 Hz, 2H), 2.23 (s, 3H), 2.19 (s, 4H), 2.14 (s, 5H), 2.11 (s, 3H), 2.03–1.78 (m, 2H), 1.63–1.49 (m, 2H), 1.48–1.37 (m, 1H). ¹³C NMR (101 MHz, CDCl₃) δ 164.0, 163.8, 157.0, 148.3, 148.2, 126.1, 125.5, 125.0, 124.8, 63.6, 62.6, 59.9, 59.8, 59.7, 54.9, 43.1, 30.2, 22.6, 13.2, 10.9. HRMS (ESI-MS) calcd. for C₂₄H₃₇N₄O₂ [M+H]⁺: 413.2911, found: 413.2905.

C1: Mn(CF₃SO₃)₂ (0.25 mmol, 1 eq.) was added to a stirred solution of chiral ligand L1 (0.25 mmol, 1 eq.) in acetonitrile (3 mL) at room temperature. The reaction mixture was stirred for 24 h. After drying under vacuum, the resulting solid was washed thoroughly with ether three times. It was then dried under vacuum to yield C1 (yield, > 90%). HRMS (ESI-MS) calcd. for C₂₃H₃₄F₃MnN₆O₃S [M-OTf]⁺: 586.1740, found: 586.1763.

C2 was prepared using a method analogous to that used for C1 starting with L2 and Mn(CF₃SO₃)₂ to obtain the product as a solid (yield, > 90%). HRMS (ESI-MS) calcd. for C₂₉H₃₈F₃MnN₆O₃S [M-OTf]⁺: 662.2058, found: 662.2070.

C3 was prepared using a method analogous to that used for C1 starting with L3 and Mn(CF₃SO₃)₂ to obtain the product as a solid (yield, > 90%). HRMS (ESI-MS) calcd. for C₂₄H₃₆MnN₄O₂ [M-2OTf]²⁺: 233.6104, found: 233.6096.

In a typical reaction, a MeCN (0.5 mL) solution consisting of the substrate (0.4 mmol), catalyst (0.2 mol%), and acid (0.5 eq.) was mixed in a 10-mL flask at −30 ℃. Then a H₂O₂ solution (1.5 eq., diluted from a 30% aqueous solution in 0.5 mL of MeCN) was added via a syringe pump over 30 min with stirring at −30 ℃. The solution was further stirred at –30 ℃ for 30 min. At this point, decane was added to the mixture as an internal reference. The reaction was quenched with saturated NaHCO₃ aqueous solution and saturated Na₂S₂O₃ aqueous solution and extracted with DCM; the sample was then investigated using GC and GC–MS analysis or purified by chromatography on silica gel to afford the epoxide product.

The structure of a ligand is known to have an important effect on the activity of metal complex catalysts. In 2013, Costas and coworkers [12, 34] reported that the electronic properties of a metal center could be tuned by introducing different substituents on the pyridine ring at the 4-position. In particular, a metal complex bearing dimethylamino groups on the ligand framework exhibited excellent activity (Scheme 1, ^Me2NPDP ligand). Further, Que and coworkers [35] captured a key oxoiron(Ⅴ) intermediate with a 50% yield, which was considered to be the active species in the oxidation reaction and was hard to synthesize and characterize, by introducing 4-methoxy-3, 5-dimethylpyridine (^dMM-pyridine) instead of pyridine donors into the tris(pyridine-2-ylmethyl)amine ligand. Recently, Wang and coworkers [36] also demonstrated, using density functional theory calculations, that an Fe(Ⅳ)-oxo cation radical species was generated when dimethylamino groups were introduced into PDP-Fe complex catalysts. These results indicated that using different substituents on ligands did change the spin state of the metal center and the activity of these metal complexes. Thus, we tried to synthesize three different tetradentate nitrogen ligands bearing strong electron-donating substituents (the Me₂N or dMM group) and the corresponding manganese complexes.

The aminopyridine ligands L1–L3 derived from L-proline were readily prepared, as shown in Scheme 2, and the manganese complexes were synthesized using the modified N4 ligands L1–L3 with Mn(OTf)₂ and stirring in CH₃CN at room temperature for 24 h.

The activities of three structurally new catalysts were then examined in asymmetric epoxidation. We selected cis-β-methylstyrene as a model substrate and used 2, 2-dimethylbutanoic acid (DMBA) as the additive [26]. Epoxidation was performed in CH₃CN at –30 ℃ using 30% aqueous H₂O₂ (1.5 eq.) as the oxidant. It was shown that the catalyst C1 (0.1 mol%) was the most effective, furnishing epoxide in 84% yield and 82% ee (Table 1, entry 1). Surprisingly, the manganese catalyst C2, with a larger steric hindrance, exhibited decreased performance despite its richer electronic properties (Table 1, entry 2). In addition, the manganese complex C3, with fewer electron-donating dMM groups, exhibited lower yield and enantioselectivity under the same conditions (Table 1, entry 3). To improve the conversion in epoxidation, a larger catalyst loading (0.2 mol%) was tested, and it was found that the substrate converted completely, and a 92% yield and 88% ee were obtained (Table 1, entry 4). The reaction proceeded well even in the presence of 50 mol% DMBA when 0.2 mol% manganese complex C1 was used (Table 1, entry 5).

In addition, the carboxylic acids also played an important role in the non-heme-metal-complex-promoted epoxidation by H₂O₂ because these reactions were achieved by the proposed "carboxylic-acid-assisted" O–O bond cleavage mechanism (Scheme 3) [16]. In 2016, Talsi and coworkers [20] studied the effect of carboxylic acids on the electronic structure of the intermediates in asymmetric epoxidation. They found that these acids with tertiary α-carbon atoms displayed a smaller g-factor anisotropy, which resulted in higher enantioselectivities [20]. Very recently, we demonstrated that only the carboxylic acid binding on the manganese center can activate H₂O₂ to produce a presumed carboxylic Mn^V=O (Ⅱ) as the oxidizing reagent (Scheme 3). Thus, the carboxylic ligand coordinated on the manganese center could affect the stereocontrol of the asymmetric epoxidation [37]. In this context, we then evaluated the contribution of different carboxylic acids to this process. The results showed that DMBA was the best additive, giving a 90% yield and 89% ee (Table 1, entry 5). Other acids, such as D-camphoric acid (D-CPA), acetic acid, 4-methylhexanoic acid (4-MHA), pivalic acid (PVA), and 2-ethylhexanoic acid (EHA), resulted in lower ee values (Table 1, entries 6–10). Further lowering the acid loading degraded the performance (Table 1, entries 11 and 12).

After establishing the optimized conditions using the manganese complex C1, we investigated the generality of the reaction of different olefins. First, various styrenes were examined under the optimal conditions. Gratifyingly, styrenes bearing different substituents were transformed to epoxides successfully, albeit with moderate enantioselectivities (Table 2, entries 1–7). In addition, the epoxidation of stilbene compounds provided epoxides in low yield, but the enantioselectivity was quite good (Table 2, entries 6 and 7). To our delight, chromenes (Table 2, entries 8 and 9) and cinnamamides (Table 2, entries 10–15) were well-tolerated, and the target products were assembled with excellent enantioselectivities (up to 99% ee).

In conclusion, we developed a series of novel tetradentate nitrogen ligands derived from L-proline. The corresponding manganese complexes were then prepared, and their catalytic activities and selectivities were evaluated in asymmetric epoxidation of olefins. Various olefins were transformed into epoxides successfully with low loading of the manganese catalyst together with aqueous H₂O₂ as the oxidant in the presence of a small amount of carboxylic acid as an additive. Notably, the enantioselective epoxidation of chromenes and cinnamamides provided the target products with excellent enantioselectivities (up to 99% ee). Unlike the substituted PDP ligands, these N4 ligands were prepared from readily available and inexpensive amino acids, thus making the manganese catalysts more practical. Further utilization of these biomimetic manganese complexes in oxidation reactions is ongoing in our laboratory.

Acknowledgments

Technical assistance, material support, and other help or advice may be acknowledged briefly in this section (excluding financial support, which should appear in the footnote on the title page).