Synthesis of Chiral Allylic Esters by Using the New Recyclable Chiral Heterogeneous Oxazoline-Based Catalysts.

A new class of recyclable supported chiral heterogeneous ligands has been synthesized by the reaction of functionalized mesoporous SBA-15 with aliphatic- and aromatic-substituted chiral amino oxazoline ligands. The obtained chiral heterogeneous ligands were characterized by several techniques such as Fourier transform infrared, X-ray diffraction, thermogravimetric analysis, scanning electron microscopy, energy-dispersive X-ray spectroscopy, and BET-BJH. The application of these new heterogeneous ligands in copper-catalyzed asymmetric allylic oxidation of olefins by using perester showed better yields up to 95% and better enantioselectivities up to 96% compared to the corresponding homogeneous catalysts. These findings can be considered as an important step in the advancement of green chemistry. Investigation of the recyclability of the catalysts confirmed that they were easily recovered and reused eight times without significant losses in reactivity, yield, and enantioselectivity.

Introduction

The formation of C–C and C–X bonds in organic synthesis is one of the most straightforward methods for the preparation of organic compounds that can proceed in the presence of homogeneous or heterogeneous catalysts.[1,2]

Despite high selectivity of chiral homogeneous catalysts, heterogeneous catalysts have attracted enormous interest because a trace amount of catalyst can be easily separated from the final products. Besides, heterogeneous catalysts are stable under different conditions, and the metal pollution of products is markedly reduced in consistent with the goals of green chemistry.[1,2] One way to prepare chiral heterogeneous catalysts is the immobilization of chiral ligands on inorganic supports such as mesoporous silica SBA-15.[3−6] In addition to having some special characteristics such as surface extension, uniform distribution, and adjustable pore diameter, SBA-15 has more advantages in comparison to other organic supports such as polymers, in which chemical stability in the reaction and mechanical resistance against the swelling in the solvent can be mentioned as examples.[7−9]

Oxazolines are an important class of chiral organic ligands.[10−14] In recent years, electron-rich chiral oxazoline ligands synthesized from inexpensive and available amino acids, having stability against hydrolysis and oxidation, produce chiral products with high stereoselectivity in catalytic asymmetric reactions.[2,11−13,15,16] The synthesis of allylic esters in the presence of copper salts is a classical method (Kharasch–Sosnovsky reaction in 1958).[17−22] Although there are several reports for the synthesis of allylic esters in the presence of transition metals such as palladium, rhodium, and iron complexes, copper complexes are more considered because of nontoxicity, inexpensiveness, and color variation by changing the oxidation states.[17−20] Asymmetric allylic oxidation of olefins without changing the double bond, producing an important intermediary such as chiral allylic ester, has gained considerable interest in organic chemistry.[23−29] In 2000, Andrus group reported the synthesis of the chiral allylic esters from cyclohexene and cyclopentene by using tert-butyl o-iodo benzoperoxoate in the presence of chiral bi-o-tolyl bisoxazoline ligands and Cu(CH3CN)4PF6, respectively, up to 68 and 61% yields and 70 and 65% enantiomeric excesses at −20 °C after 120 h in acetonitrile.[30] In our previous work, allylic oxidation of cyclic olefins was reported in the presence of copper(I) complexes of the immobilized chiral 4-oxazolinylaniline ligands on mesoporous silicas MCM-41, which produce the products with moderate to good enantiomeric excess at 0 °C. However, long time of reaction, a high ratio of olefin to perester, a large amount of required catalyst, and harsh conditions for the preparation of oxazolinylaniline ligands were needed.[11] For our laboratory purposes, in order to improve the previously achieved results, increase in the efficiency of the reaction and development in the diversity of chiral allylic esters, enantioselective allylic sp3 C–H oxidation of olefins was performed in the presence of chiral heterogeneous copper complexes which was obtained via in situ immobilization of grafted chiral amino oxazoline ligands by various copper salts.

Results and Discussion

Catalyst Characterization

In the Fourier transform infrared (FT-IR) spectrum of mesoporous silica SBA-15, a broad band of SiO–H in around ν̅ 3200–3600 cm–1 and the asymmetric stretching and symmetrical vibration of Si–O–Si in ν̅ 1000–1200 and 807 cm–1 were observed. In the case of Cl-SBA-15, the observation of the OH, CH2, and C–Cl groups in ν̅ 3423, 2958, and 788 cm–1 clearly demonstrated the immobilization of 3-chloropropyltrimethoxysilan on SBA-15. In the FT-IR spectra of the chiral ligands OX-R-SBA-15, related peaks to the sp2 C–H, sp3 C–H, and C=N bonds in ν̅ 3080–3082, 2956–2962, and 1612–1639, respectively, proved the presence of oxazoline groups[3,6,11] (Figure ).

Figure 1

FT-IR spectra of mesoporous silica SBA-15, Cl-SBA-15, and OX-R-SBA-15.

As shown in the thermogravimetric analysis (TGA) (Figure ), the weight loss of 2–4 wt % in the temperature range 0–150 °C corresponds to the elimination of water that is physically and chemically adsorbed on the surfaces. The weight loss in the temperature range 150–500 °C was 10.39% for Cl-SBA-15 that shows the loading of 3-chloropropyltrimethoxysilane (CPTMS) on SBA-15, and weight losses for chiral heterogeneous ligands OX-R-SBA-15, due to the decomposition of chiral amino oxazolines from the surface of SBA-15, were 12.7–14.7% (0.18–0.27 mmol oxazoline/g SBA-15).[3,6,11,31]

Figure 2

TGA of Cl-SBA-15 and OX-R-SBA-15.

According to the X-ray diffraction (XRD) pattern of the calcined SBA-15, a very strong peak (100) and two reflections (110) and (200) with the lower intensity were observed. The XRD patterns of Cl-SBA-15 and OX-R-SBA-15 showed similar features as calcined mesoporous silica SBA-15 with the lower intensity of (100) which shifted from 2θ = 0.79 to 1.13° because of functionalization of the organic group. It should be noted that the shifting of 2θ for catalyst with aromatic substituents on amino oxazoline ligands is more than that of aliphatic ones. Moreover, in the functionalization process, d100 and a0 were decreased from 11.18 to 7.81 and 12.90 to 9.02 nm, respectively. These results were confirmed that the mesoporous structure of SBA-15 remained almost intact during the functionalization, and the hexagonally arranged pore arrays were only decreased[3,11,31] (Figure ).

Figure 3

XRD patterns of mesoporous silica SBA-15, Cl-SBA-15, and OX-R-SBA-15.

The surface morphology of the SBA-15, Cl-SBA-15, and OX-R-SBA-15 were investigated by scanning electron microscopy (SEM) (Figure ). They revealed that mesoporous silica SBA-15 has a shape close to the short rod with the friable surface. The shape of the resulting inorganic–organic hybrid materials was also short rodlike after grafting on SBA-15 material.[31−33] These results showed that the chiral amino oxazoline group was successfully immobilized on Cl-SBA-15 nanoparticles without considerable changes during the immobilization process. The energy-dispersive X-ray spectroscopy (EDX) was also performed to prove the presence of the elements of silicium, oxygen, nitrogen, and carbon in the synthesized chiral heterogeneous ligand OX-Bn-SBA-15 (Figure ).

Figure 4

Scanning electron micrographs of (a) mesoporous silica SBA-15, (b) Cl-SBA-15, and (c) OX-Bn-SBA-15 at 1 μm scale and (d) OX-Bn-SBA-15 at 2 μm scale.

Figure 5

EDX spectrum of OX-Bn-SBA-15.

To elucidate the effect of the immobilization of the chiral amino oxazoline ligand on the textural features, the N2 adsorption–desorption isotherms of SBA-15, Cl-SBA-15, and OX-Bn-SBA-15 were recorded. Based on the Table , BET surface area, pore volume, and average pore diameter of mesoporous silica SBA-15 decreased by immobilization of 3-chloropropyltrimethoxysilan in Cl-SBA-15 and chiral amino oxazoline in OX-Bn-SBA-15. The results demonstrated the successful immobilization of chiral ligands inside the mesopore channels, resulting in a partial blockage of the mesopore channels. In addition, with increasing organic group content in Cl-SBA-15 and OX-Bn-SBA-15, the surface area, pore volume, and average pore diameter of SBA-15 more severely reduced.[33,34]

Table 1

N2 Adsorption–Desorption Information of Mesoporous Silica SBA-15, Cl-SBA-15, and OX-Bn-SBA-15

sample	d100 (nm)a	a0 (nm)b	BET surface area (m2/g)	BET total pore volume (cm3/g)	average pore diameter (nm)	BJH pore volume (cm3/g)	WTH (nm)c
SBA-15	11.17	12.90	845	1.22	12.56	1.20	0.34
Cl-SBA-15	9.76	11.27	503	0.96	10.60	0.904	0.67
OX-Bn-SBA-15	7.88	9.10	345	0.63	8.10	0.601	1.00

λ = 2d100 sin θ (λ = 1.54060 Cu).

Unit cell parameter: a0 = 2d100/√3.

Wall thicknesses were calculated as: a0 – pore size.

As outlined in Figure , based on the N2 adsorption–desorption isotherms of the functionalized mesoporous silica, the isotherm of OX-Bn-SBA-15 (6b) is classified as type IV characteristic with an H1 hysteresis loop which is typically observed in mesoporous materials with a two-dimensional cylindrical channel.[33,34] In the first part of the isotherm of OX-Bn-SBA-15 at low pressure (p/p0 = 0–0.5), micropores with pores smaller than 2 nm is saturated with nitrogen gas; therefore, the smallness of this part represents the small number of micropores. The observed sharp inflection in the p/p0 range of 0.65 to 0.78 of isotherm is related to the characteristic of capillary condensation within a uniform mesostructure (Figure a). Pore size distribution curve shows a narrow distribution for OX-Bn-SBA-15 (Figure b).

Figure 6

(a) N2 adsorption–desorption isotherm of OX-Bn-SBA-15; (b) pore size distribution curve.

Catalytic Activity

There are some problems in asymmetric allylic oxidation: low activity of catalysts, long time of reaction, and moderate enantioselectivity. To overcome these problems, we focused on the synthesis of heterogeneous catalysts with high activity and enantioselectivity.[35] For this purpose, we prepared aliphatic- and aromatic-substituted amino oxazoline ligands 3a–d(14) in two steps: first, chiral amino alcohols 2a–d were synthesized by reduction of amino acids 1a–d with NaBH4/I2 in dried tetrahydrofuran;[36] next, they were converted to chiral ligands 3a–d in the presence of cyanogen bromide (CNBr)[37] in dried tetrahydrofuran with good yields (Scheme ). Consequently, the chiral heterogeneous ligands OX-R-SBA-15 (4a–d) were obtained by grafting the amino oxazoline ligands 3a–d on Cl-SBA-15 which have been synthesized by functionalization of SBA-15 with CPTMS.[3,38]

Scheme 1

Synthesis of Chiral Amino Oxazoline Ligands 3a–d and Immobilization on Cl-SBA-15

Chiral heterogeneous catalysts were obtained by in situ immobilization of copper salts and applied in the asymmetric allylic oxidation of alkenes using various substituted peresters (Scheme ). In this work, the optimal condition was obtained by changing many parameters such as type and amount of copper salt and chiral heterogeneous ligands, solvent, and temperature. The model reaction was carried out by using cyclohexene (3 mmol) and tert-butyl o-iodoperbenzoate (0.85 mmol) in the presence of [Cu(CH3CN)4]PF6[39] (3.5 mol %) and OX-Bn-SBA-15 4b (10 mg) in acetonitrile at room temperature. Initially, we investigated the effect of various copper salts such as Cu(OAc)2, CuSO4, Cu(NO3)2, CuCl2, CuO, Cu2O, CuI, CuOTf, Cu(OTf)2, and [Cu(CH3CN)4]PF6. As resulted, the copper(I) complexes gave more enantioselectivity and reactivity in comparison with copper(II) at room temperature and [Cu (CH3CN)4]PF6 was the best copper source (entry 10). We next examined the effect of different solvents on enantioselectivity and reactivity (entries 10–15). Although the initial rate of the reaction was fast in CH2Cl2 and CHCl3, but ee value, yield, and reaction conversion were improved in acetonitrile. In toluene, the rate of the reaction was slow, and because of the presence of the benzylic C–H bond, a side product was observed.

Scheme 2

Preparation of Chiral Allylic Esters in the Presence of Synthesized Chiral Heterogeneous Catalysts

The effect of ligand substitutions was also evaluated on the reaction, and the result indicated the ligands with aromatic groups on the oxazoline rings (4a, 4b) led to the allylic esters in more enantioselectivity and reactivity in comparison to the ligands with aliphatic substitutions (4c, 4d). It might be that the aryl substituents play an important role in ligand efficiency because of the interaction between allyl radicals and aryl substituents at the transition states which is known to be the reason for such an effect.[40,41] It should be mentioned that the yield of products in the presence of aromatic and aliphatic substitutions was approximately equaled. Based on the Table results, OX-Bn-SBA-15 4b was chosen as the best heterogeneous ligand in the optimal reaction (entry 10 vs 16–18). By altering the amounts of [Cu(CH3CN)4]PF6 and OX-Bn-SBA-15, the best combination found was 10 mg of [Cu(CH3CN)4]PF6 (3.5 mol %) and 10 mg of OX-Bn-SBA-15 (4b) (entry 10 vs 23–26). By lowering the reaction temperature, the reaction rate was decelerated, but the enantiomeric excess slightly increased. However, the enantioselectivity and yield are aligned at the room temperature. On the basis of a comparison of the optical rotations with literature values, the absolute configuration of the products was assigned. It should be noted that the stereochemistry of products in this work depends on the absolute configuration of ligands, so that prepared ligands from S-amino alcohols induce S configuration in the resulted allylic esters, and this also happens for R configuration. Ligands ent-4a and ent-4b with R stereochemistry unlike other ligands 4a–d transferred R configuration to products (entries 33–34, Table ). The experiments were controlled by using the ligand 3b and N-methyl protected ligand 3b () as homogeneous catalysts. The obtained results show that induced chirality better than 3b. However, their catalytic activities are generally lower than heterogeneous catalysts (entries 35–36, Table ). On the other hand, when the reaction was performed in the presence of isolated chiral nanocatalysts Cu-4 that were prepared by immobilizing copper salts on the chiral heterogeneous ligand, no significant progress was observed.

Table 3

Synthesis of Chiral Allylic Esters from Different Alkenes by Using Various Perestersa

Reaction conditions: alkene (3 mmol), perester 7 (0.85 mmol), OX-Bn-SBA-15 (6b) (10 mg), Cu(CH3CN)4PF6 (3.5 mol %), CH3CN (2 mL) at room temperature.

Isolated yield based on perester.

Enantiomeric excess (ee) were determined by HPLC on Chiralpak AD and/or Chiralcel OD-H and/or Nucleocel Alpha S columns.

Andrus group result in 2000.[30]

Our pervious result in 2016.[11]

Table 2

Effects of Copper Salt, Chiral Heterogeneous Ligand, Solvent, Temperature, and Cu(CH3CN)4PF6 Loading and Chiral Heterogeneous Ligand Loading in the Allylic C–H Bonds Oxidation of Cyclohexenea

entry	Cu salt (mol %)	chiral ligand (mg)	solvent (2 mL)	temperature (°C)	time (h)	yield (%)b	ee (%)c
1	Cu(OAc)2 (3.5)	6b (10)	acetonitrile	r.t.	85	60	35
2	CuSO4 (3.5)	6b (10)	acetonitrile	r.t.	65	63	41
3	Cu(NO3)2 (3.5)	6b (10)	acetonitrile	r.t.	70	68	32
4	CuCl2 (3.5)	6b (10)	acetonitrile	r.t.	65	62	40
5	CuO (3.5)	6b (10)	acetonitrile	r.t.	61	73	48
6	Cu2O (3.5)	6b (10)	acetonitrile	r.t.	53	61	42
7	CuI (3.5)	6b (10)	acetonitrile	r.t.	47	80	57
8	Cu(OTf)2 (3.5)	6b (10)	acetonitrile	r.t.	45	75	71
9	Cu(OTf) (3.5)	6b (10)	acetonitrile	r.t.	35	87	75
10	Cu(CH3CN)4PF6(3.5)	6b (10)	acetonitrile	r.t.	24	90	90
11	Cu(CH3CN)4PF6 (3.5)	6b (10)	acetone	r.t.	55	65	68
12	Cu(CH3CN)4PF6 (3.5)	6b (10)	dichloromethane	r.t.	50	45	38
13	Cu(CH3CN)4PF6 (3.5)	6b (10)	chloroform	r.t.	45	61	30
14	Cu(CH3CN)4PF6 (3.5)	6b (10)	toluene	r.t.	72	35	24
15	Cu(CH3CN)4PF6 (3.5)	6b (10)	tetrahydrofuran	r.t.	55	47	32
16	Cu(CH3CN)4PF6 (3.5)	6a (10)	acetonitrile	r.t.	28	95	81
17	Cu(CH3CN)4PF6 (3.5)	6c (10)	acetonitrile	r.t.	45	87	41
18	Cu(CH3CN)4PF6 (3.5)	6d (10)	acetonitrile	r.t.	49	82	46
19	Cu(CH3CN)4PF6 (7.0)	6b (10)	acetonitrile	r.t.	23	90	83
20	Cu(CH3CN)4PF6 (1.75)	6b (10)	acetonitrile	r.t.	45	75	78
21	Cu(CH3CN)4PF6 (0.9)	6b (10)	acetonitrile	r.t.	67	55	75
22	Cu(CH3CN)4PF6 (0.4)	6b (10)	acetonitrile	r.t.	81	50	65
23	Cu(CH3CN)4PF6 (3.5)	6b (20)	acetonitrile	r.t.	37	83	96
24	Cu(CH3CN)4PF6 (3.5)	6b (5)	acetonitrile	r.t.	50	77	75
25	Cu(CH3CN)4PF6 (3.5)	6b (2.5)	acetonitrile	r.t.	51	75	58
26	Cu(CH3CN)4PF6 (3.5)	6b (1)	acetonitrile	r.t.	50	35	37
27	Cu(CH3CN)4PF6 (3.5)	6b (10)	acetonitrile	10	30	80	85
28	Cu(CH3CN)4PF6 (3.5)	6b (10)	acetonitrile	5	45	73	83
29	Cu(CH3CN)4PF6 (3.5)	6b (10)	acetonitrile	0	72	68	93
30	Cu(CH3CN)4PF6 (3.5)	6b (10)	acetonitrile	–5	90	62	95
31	Cu(CH3CN)4PF6 (3.5)	6b (10)	acetonitrile	–10	102	54	95
32	Cu(CH3CN)4PF6 (3.5)	6a (10)	acetonitrile	–16	112	45	90
33d	Cu(CH3CN)4PF6 (3.5)	ent-6b (10)	acetonitrile	r.t.	32	85	–80
34d	Cu(CH3CN)4PF6 (3.5)	ent-6a (10)	acetonitrile	r.t.	28	90	–70
35e	Cu(CH3CN)4PF6 (3.5)	3b (10)	acetonitrile	r.t.	72	30	15
36e	Cu(CH3CN)4PF6 (3.5)	N–Me-3b (10)	acetonitrile	r.t.	45	70	50

Reaction conditions: cyclohexene (3 mmol), perester 7b (0.85 mmol), chiral ligand (mg), Cu Salt (mol %), and solvent (2 mL).

Isolated yield based on perester.

The enantiomeric excess (ee) was determined by HPLC with Chiralpak AD column; eluent: n-hexane/isopropyl alcohol = 99.6/0.4; flow rate: 0.6 mL/min; tR = 22.3 min (R), 24.9 min (S).

The configuration of chiral ligands ent-6b and ent-6a were R.

Homogeneous catalysts.

The obtained optimized condition was investigated using different peresters such as tert-butyl benzoperoxoate 3a, tert-butyl-o-iodoperbenzoate 3b, tert-butyl-o-bromo perbenzoate 3c, tert-butyl-o-chloroperbenzoate 3d, tert-butyl-p-nitrobenzoperoxoate 3e, tert-butyl-p-chlorobenzoperoxoate 3f, and tert-butyl-p-methylbenzoperoxoate 3g (Table ). The peresters 3b and 3e gave the products with better enantioselectivities and yields than others. In addition, we investigated various cycloalkenes under optimized condition, and chiral allylic esters were obtained with better enantiomeric excesses and yields in comparison of previous reports.[11,30] When the reaction was carried out using 1,5-cyclooctadiene, the product was achieved in a shorter time with good enantiomeric excess and excellent yield (Table ). It could be because of the twist-boat conformation of 1,5-cyclooctadiene[42−44] which can be fitted to the catalyst complex cavity. For acyclic substrates such as 1-hexene, 1-octene, and ally benzene, the reaction proceeded with low yields up to 23% and low enantioselectivities up to 18%. The remarkable point was that by decreasing temperature for acyclic substrates, the reaction did not progress. To evaluate the efficiency of this protocol, the catalytic activity of the OX-R-SBA-15 was compared with immobilized chiral 4-oxazolinylaniline ligands on mesoporous silica MCM-41.[11] This comparison proved that a low amount of OX-R-SBA-15 was catalyzed, the reaction with remarkably higher reaction rate, yield and enantiomeric excess (entries 7–10 in Table ).

Catalyst Recyclability

The recyclability of the synthesized chiral heterogeneous ligands was investigated in the asymmetric allylic oxidation of olefins. For this aim, the catalyst was isolated at the end of the reaction by simple filtration, washed with CH3OH/CH2Cl2, and subsequently dried at 50 °C under vacuum for 3 h. The recovered catalyst was used in further reaction runs that is able to accomplish the allylic oxidation for eight consecutive catalytic runs, without a remarkable effect on the enantioselectivity, time of reaction, and chemical yield. On the other hand, ee, yield, and reactivity were clearly decreased after the eighth run. Moreover, the FT-IR spectrum and XRD pattern of heterogeneous ligand 6b remain unchanged after the eighth catalytic run, indicating that the mesoporous structure of 6b was preserved (Figure ).

Figure 7

Correlation between ee, yield, and time of reaction with recyclability of OX-Bn-SBA-15.

Conclusions

In conclusion, we synthesized a class of aromatic- and aliphatic-substituted amino oxazoline ligands from available amino acids and immobilized them on inorganic support SBA-15 using the post-grafting method for the first time. Then, the catalytic activity of the new oxazoline-based heterogeneous ligands was investigated in Cu-catalyzed asymmetric allylic oxidation of sp3 C–H bonds of cyclic and acyclic olefins using various peresters in different conditions at a temperature range of −16 to 25 °C. As it can be concluded from the result, in the presence of chiral heterogeneous ligands with aromatic substituents on the oxazoline ring, allylic esters were obtained in more enantioselectivities and reactivities than with aliphatic substituents. In terms of copper salts and solvents, with ligand 6b, the best enantioselectivitie was achieved using Cu(CH3CN)4PF6 in acetonitrile. Among cyclic olefins were examined, cyclohexene and 1,5-cyclooctadiene gave the best results in regard to enantioselectivity and chemical yield. Unlike cyclic olefins, acyclic olefins did not give good results in this reaction. The recyclability study of the heterogeneous ligand (OX-Bn-SBA-15) confirmed that it was recycled and recovered up to eight times without a remarkable decrease in enantioselectivity, reactivity, and yield. Further investigation on the potential applications of the synthesized chiral allylic esters containing halogens as chiral precursors in palladium-catalyzed Mizoroki–Heck coupling reaction is currently underway in our laboratory.

Experimental Section

Typical Procedure for the Preparation of Chiral Amino Oxazolines 3a–d

In an oven-dried round-bottom flask, (S)-2-amino-3-phenylpropan-1-ol (2b)[36] (3.0 mmol, 0.50 g) (Pages S2–S4 and Figures S1–S8) was dissolved in dried THF (6 mL) under a nitrogen atmosphere. Then, the resulting solution was cooled in an ice-bath, and potassium carbonate (3.6 mmol, 0.50 g) was added. Next, the mixture of the reaction was slowly charged by the solution of cyanogen bromide (3.6 mmol, 0.38 g) in dried THF (4 mL) at 0 °C. After that, the ice bath was removed, and the resulting mixture was stirred at room temperature for another 1.5 h. After evaporation of THF, extraction was performed with ethyl acetate (2 × 10 mL) and washed with brine (10 mL). The combined organic layers were dried over magnesium sulfate. The product 3b was obtained in 92% yield. Other products 3a, 3c, and 3d were also prepared from the corresponding chiral amino alcohols in the similar procedure in good yields up to 95%[14,45−55] (Figures S9–S16).

(S)-4-Phenyl-4,5-dihydrooxazol-2-amine (3a)

mp 120–123 °C; 1H NMR (400 MHz, CDCl3): δH (ppm) 4.10 (1H, t, J = 6.6 Hz, H2C–O), 4.66 (1H, t, J = 8.6 Hz, H2C–O), 5.13 (1H, t, J = 8.2 Hz, *CH), 7.28–7.38 (5H, m, Ar); 13C NMR (100 MHz, CDCl3): δC (ppm) 67.7, 75.4, 126.4, 127.4, 128.6, 143.5, 161.6; [α]D25 +5.2° (c 0.5, CHCl3).

(S)-4-Benzyl-4,5-dihydrooxazol-2-amine (3b)

mp 84–89 °C; 1H NMR (400 MHz, CDCl3): δH (ppm) 2.72 (1H, dd, J = 13.6, 7.6 Hz, H2C-Ph), 2.98 (1H, dd, J = 13.6, 6.0 Hz, H2C-Ph), 3.91–4.01 (1H, m, *CH), 4.30 (2H, dd, J = 15.2, 4.6 Hz, H2C–O), 7.31–7.36 (5H, m, Ar); 13C NMR, (100 MHz, CDCl3): δC (ppm) 42.5, 65.5, 72.7, 126.3, 128.5, 129.2, 138.5, 160.7; [α]D25 −18.5° (c 0.5, CHCl3).

(S)-4-Isopropyl-4,5-dihydrooxazol-2-amine (3c)

mp 131–134 °C; 1H NMR (400 MHz, CDCl3): δH (ppm) 0.92 (3H, d, J = 6.8 Hz, CH3), 0.98 (3H, d, J = 6.6 Hz, CH3), 1.66–1.71 (1H, m, CH), 3.75–3.81 (1H, m,*CH), 3.98 (1H, t, J = 7.6 Hz, H2C–O), 4.32 (1H, t, J = 8.4 Hz, H2C–O); 13C NMR (100 MHz, CDCl3): δC (ppm) 18.3, 18.9, 33.3, 70.1, 71.3, 160.5; [α]D25 −25.3° (c 0.5, CHCl3).

(S)-4-Isobutyl-4,5-dihydrooxazol-2-amine (3d)

mp 105–106 °C; 1H NMR (400 MHz, CDCl3): δH (ppm) 0.92 (6H, d, J = 6.4 Hz, CH3) 1.22–1.36 (1H, m, HC–CH3), 1.48–1.55 (1H, m, H2C–CH), 1.66–1.76 (1H, m, H2C–CH), 3.80 (1H, t, J = 7.4 Hz,*CH), 4.03 (1H, t, J = 7.6 Hz, H2C–O), 4.35 (1H, t, J = 8.2 Hz, H2C–O); 13C NMR (100 MHz, CDCl3): δC (ppm) 22.3, 23.1, 25.4, 45.9, 62.5, 73.7, 160.7; [α]D25 −21.6° (c 0.5, CHCl3).

Typical Procedure for the Preparation of Chiral Heterogeneous Ligands 4a–d (OX-R-SBA-15)

Under a nitrogen atmosphere, a flame-dried round-bottom flask (50 mL) was charged by 0.50 g of functionalized SBA-15 (Cl-SBA-15)[3,38] (Pages S4 and S5) and 10 mL of dried toluene. Next, 2 mmol (0.28 mL) of triethylamine and chiral amino oxazoline (S)-4-benzyl-4,5-dihydrooxazol-2-amine (3b) (2 mmol, 0.35 g) were added and refluxed for 24 h. After that, the obtained powder was separated by filtration and washed thoroughly with CH2Cl2/CH3OH (3 × 10 mL), and then Soxhlet-extracted with 30 mL of CH2Cl2/CH3OH (1:1) to remove any unreacted species. The light gray solid as chiral heterogeneous ligand 4b was dried at room temperature. The preservation of mesoporous structure of the chiral heterogeneous ligand OX-Bn-SBA-15 was proved by FT-IR, XRD, SEM, EDX, TGA-DTA, and BET–BJH techniques.[11] Other heterogeneous ligands 4a and 4c–d were synthesized in a similar procedure.

Typical Procedure for Asymmetric Allylic C–H Bond Oxidation of Olefins

To a flame-dried Schlenk tube under the nitrogen atmosphere, 10 mg of chiral heterogonous ligand 6b, [Cu(CH3CN)4]PF6 (10 mg, 0.03 mmol), and dried acetonitrile (2 mL) were added and stirred for 2 h at room temperature. Next, cyclohexene (3 mmol, 0.3 mL) and tert-butyl-2-iodobenzoperoxoate (7b) (0.85 mmol, 0.27 g) (Pages S5 and S6 and Figures S17 and S18) were slowly added to the mixture of reaction. The resulting mixture was stirred at room temperature until TLC revealed complete disappearance of 7b (Scheme ). After that, the chiral heterogeneous catalyst was separated by filtration to recycle, and acetonitrile was removed under the reduced pressure. The obtained residue extracted three times with aqueous ammonium hydroxide (10%) and ethyl acetate. The combined organic phase was washed with 10 mL of NaHCO3 (10%), dried over anhydrous MgSO4, and then purified by silica gel column chromatography using n-hexane and ethyl acetate (90:10) to provide the (S)-cyclohex-2-en-1-yl 2-iodobenzoate (8b). Other chiral allylic esters 8–14 were synthesized in a similar procedure (yield up to 95%) (Figures S19–S32).

(S)-Cyclohex-2-en-1-yl Benzoate (8a)[56]

1HNMR (400 MHz, CDCl3): δH (ppm) 1.62–2.16 (6H, m, CH2), 5.54 (1H, dd, J = 3.4, 1.6 Hz, H–C*O), 5.88–5.89 (1H, m, HC=CH), 6.03–6.05 (1H, m, HC=CH), 7.45–7.48 (2H, m, Ar), 7.56–7.60 (1H, m, Ar), 8.07–8.10 (2H, m, Ar); 13CNMR (100 MHz, CDCl3): δC (ppm) 19.0, 25.0, 28.4, 68.6, 125.8, 128.2, 129.6, 130.8, 132.8, 132.9, 166.2; m/z (%): 202 (0.5, M), 122 (23), 111 (37), 97 (54), 85 (53), 57 (100); [α]D20 −70.5° (c 1.0, CHCl3); the optical purity was determined by HPLC with Chiralcel OD-H column; eluent: n-hexane/isopropyl alcohol = 99.7/0.3; flow rate: 0.6 mL/min; tR = 16.6 min (R), 18.5 min (S) (maximum ee = 60%).

(S)-Cyclohex-2-en-1-yl 2-Iodobenzoate (8b)[30,41]

FT-IR (KBr) (νmax/cm–1): 3063, 2932, 1722, 1649, 462; 1HNMR (300 MHz, CDCl3): δH (ppm) 1.95–2.13 (6H, m, CH2), 5.54 (1H, d, J = 2.8 Hz, H–C*O), 5.90 (1H, d, J = 1.4 Hz, HC=CH), 6.00–6.05 (1H, m, HC=CH), 7.12–7.14 (1H, m, Ar), 7.40 (1H, t, J = 7.5 Hz, Ar), 7.79 (1H, d, J = 7.6 Hz, Ar), 7.99 (1H, d, J = 7.9 Hz, Ar); 13CNMR (75 MHz, CDCl3): δC (ppm) 18.8, 24.9, 29.7, 69.5, 93.9, 125.2, 127.9, 130.9, 132.4, 133.3, 135.7, 141.2, 166.4; m/z (%): 328 (0.5, M), 248 (6), 231 (11), 127 (15), 111 (25), 97 (49), 85 (71), 57 (100); Anal Calcd. for C13H13IO2: C, 47.58; H, 3.99. Found: C, 47.59; H, 3.92; [α]D20 −169.0° (c 1.0, CHCl3); the optical purity was determined by HPLC with Chiralpak AD column; eluent: n-hexane/isopropyl alcohol = 99.6/0.4; flow rate: 0.6 mL/min; tR = 22.3 min (R), 24.9 min (S) (maximum ee = 96%).

(S)-Cyclohex-2-en-1-yl 4-Nitrobenzoate (8e)[14,41,56,57]

mp 69–70 °C (lit. 68–71 °C); [α]D20 −133.7° (c 1.0, CHCl3); the optical purity was determined by HPLC with Nucleocel Alpha S column; eluent: n-hexane/isopropyl alcohol = 99.5/0.5; flow rate: 0.5 mL/min; tR = 29.0 min (R), 31.5 min (S) (maximum ee = 90%).

(S)-Cyclohex-2-en-1-yl 4-Chlorobenzoate (8f)[56]

FT-IR (KBr, cm–1): 3070, 2937, 1720, 1652, 1452, 1053, 480; 1HNMR (400 MHz, CDCl3): δH (ppm) 1.71–2.20 (6H, m, CH2), 5.53 (1H, d, J = 2.0 Hz, H–C*O), 5.84 (1H, dd, J = 10.0, 1.6 Hz, HC=CH), 6.02–6.06 (1H, m, HC=CH), 7.43 (2H, d, J = 8.4 Hz, Ar), 8.01 (2H, d, J = 8.4 Hz, Ar); 13CNMR (100 MHz, CDCl3): δC (ppm) 18.9, 24.9, 28.4, 68.9, 125.5, 128.6, 129.3, 131.0, 133.1, 139.2, 165.4; [α]D20 −137.0° (c 1.0, CHCl3); the optical purity was determined by HPLC with Chiralpak AD column; eluent: n-hexane/isopropyl alcohol = 99.7/0.3; flow rate: 0.6 mL/min; tR = 22.8 min (R), 24.3 min (S) (maximum ee = 76%).

(S)-Cyclopent-2-en-1-yl 2-Iodobenzoate (9b)[30]

FT-IR (KBr, cm–1): 3054, 2926, 1718, 1653, 458; 1HNMR (300 MHz, CDCl3): δH (ppm) 2.38–2.43 (3H, m, CH2), 2.57–2.60 (1H, m, CH2), 5.96–6.00 (2H, m, HC=CH and H-*C–O), 6.19–6.21 (1H, m, HC=CH), 7.14–7.17 (1H, m, Ar), 7.39 (1H, t, J = 7.2 Hz, Ar), 7.75–7.80 (1H, m, Ar), 7.98 (1H, d, J = 7.9 Hz, Ar); 13C NMR (75 MHz, CDCl3): δC (ppm) 29.8, 31.2, 82.1, 93.9, 127.9, 128.9, 130.9, 132.4, 135.7, 138.3, 141.2, 166.7; m/z (%): 314 (0.3, M), 248 (9), 127 (27), 111 (38), 97 (65), 85 (100), 57 (78); Anal. Calcd for C12H11IO2: C, 45.88; H, 3.53. Found: C, 45.90; H, 3.44; [α]D20 −97.5° (c 1.0, CHCl3); the optical purity was determined by HPLC with Chiralpak AD column; eluent: n-hexane/isopropyl alcohol = 99.6/0.4; flow rate: 0.6 mL/min; tR = 20.5 min (R), 22.3 min (S) (maximum ee = 87%).

(S)-Cyclopent-2-en-1-yl 4-Nitrobenzoate (9e)[14,56,57]

mp 75–77 °C (lit. 77–79 °C); [α]D20 −161.2° (c 1.0, CHCl3); the optical purity was determined by HPLC with Nucleocel Alpha S column; eluent: n-hexane/isopropyl alcohol = 99.5/0.5; flow rate: 0.4 mL/min; tR = 35.1 min (R), 36.7 min (S) (maximum ee = 86%).

(S)-Cyclooct-2-en-1-yl 2-Iodobenzoate (10b)[41]

FT-IR (KBr, cm–1): 3072, 2967, 1717, 1644, 475; 1H NMR (300 MHz, CDCl3): δH (ppm) 1.61–1.66 (4H, m, CH2), 2.13–2.17 (4H, m, CH2), 2.27–2.40 (1H, m,, CH2), 2.88–2.92 (1H, m, CH2), 5.62–5.72 (2H, m, HC=CH and H–C–O), 5.90–5.95 (1H, m, HC=CH), 7.12–7.17 (1H, m, Ar), 7.40 (1H, t, J = 7.6 Hz, Ar), 7.78–7.81 (1H, m, Ar) 7.98 (1H, d, J = 7.9 Hz, Ar); 13CNMR (75 MHz, CDCl3): δC (ppm) 23.4, 25.8, 26.3, 28.8, 35.1, 74.1, 94.1, 127.9, 130.3, 130.6, 131.1, 132.5, 135.6, 141.2, 166.0; m/z (%): 356 (0.4, M), 248 (12), 230 (6), 127 (28), 124 (100), 57 (65); Anal. Calcd for C15H17IO2: C, 50.58; H, 4.81. Found: C, 50.57; H, 4.84; [α]D20 +57.6° (c 1.0, CHCl3); the optical purity was determined by HPLC with Chiralpak AD column; eluent: n-hexane/isopropyl alcohol = 99.6/0.4; flow rate: 0.6 mL/min; tR = 16.5 min (R), 19.1 min (S) (maximum ee = 85%).

(S)-Cyclooct-2-en-1-yl 4-Nitrobenzoate (10e)[14,41,58]

mp 71–73 °C (lit. 71–74 °C); [α]D20 +37.2° (c 1.0, CHCl3); the optical purity was determined by HPLC with Chiralcel OD-H column; eluent: n-hexane/isopropyl alcohol = 99.5/0.5; flow rate: 0.4 mL/min; tR = 18.3 min (R), 21.8 min (S) (maximum ee = 72%).

(S)-Cycloocta-2,6-dien-1-yl 2-Iodobenzoate (11b)

FT-IR (KBr, cm–1): 3080, 2959, 1737, 1659, 486; 1H NMR (300 MHz, CDCl3): δH (ppm) 2.15–2.35 (2H, m, CH2), 2.53–2.60 (3H, m, CH2), 2.83–2.88 (1H, m, CH2), 5.62–5.68 (4H, m, HC=CH and H–C–O), 6.23–6.25 (1H, m, HC=CH), 7.13 (1H, d, J = 6.2 Hz, Ar), 7.36–7.40 (1H, m, Ar), 7.78 (1H, d, J = 6.4 Hz, Ar), 7.96 (1H, d, J = 7.3 Hz, Ar); 13CNMR (75 MHz, CDCl3): δC (ppm) =27.9, 28.0, 33.9, 74.0, 94.1, 125.1, 127.9, 128.7, 129.6, 129.8, 130.9, 132.6, 135.4, 141.2, 166.0; m/z (%): 354 (0.7, M), 248 (34), 228 (10), 127 (19), 122 (100), 57 (65), 54 (100); Anal. Calcd for C15H15IO2: C, 50.87; H, 4.27. Found: C, 50.88; H, 4.22; [α]D20 +31.0° (c 1.0, CHCl3); the optical purity was determined by HPLC with Chiralcel OD-H column; eluent: n-hexane/isopropyl alcohol = 99.6/0.4; flow rate: 0.6 mL/min; tR = 28.3 min (S), 30.4 min (R) (maximum ee = 88%).

(S)-Cycloocta-2,6-dien-1-yl 4-Nitrobenzoate (11e)[14,41,56−58]

mp 73–75 °C (lit. 74–76 °C) [α]D20 +23.8° (c 1.0, CHCl3); the optical purity was determined by HPLC with Chiralcel OD-H column; eluent: n-hexane/isopropyl alcohol = 99.3/0.7; flow rate: 0.4 mL/min; tR = 31.5 min (S), 33.7 min (R) (maximum ee = 92%).