Asymmetric Epoxidation of 1,1-Disubstituted Terminal Olefins by Chiral Dioxirane via a Planar-like Transition State

Abstract

Various 1,1-disubstituted terminal olefins have been investigated for asymmetric epoxidation using chiral ketone catalysts. Up to 88% ee has been achieved with a lactam ketone, and a planar transition state is likely to be a major reaction pathway.

Introduction

Chiral dioxiranes have recently been shown to be effective for asymmetric epoxidation of olefins, and a number of laboratories have extensively investigated chiral ketones of various structures.¹ In our own studies, we found that fructose-derived ketone 1 is a very effective catalyst for the epoxidation of trans- and trisubstituted olefins,² and oxazolidinone-bearing ketones 2 can give high ee’s for olefins such as conjugated aromatic cis-olefins,3a,c,d,e,g conjugated cis-dienes^3k and enynes,3a,c,l styrenes,3b,c,d,f certain trisubstituted,3h,j and tetrasubstituted olefins^3i,j which had not been effective with ketone 1 (Figure 1). Studies have shown that the enantioselectivity afforded by ketone 2 results from an apparent attractive interaction between the R_π group of the olefin and the spiro oxazolidinone of the ketone catalyst.³

Figure 1.

Among six classes of olefins (Figure 2), 1,1-disubstituted terminal olefins (VI) have generally been challenging for asymmetric epoxidation.^4–8 Epoxidation of α-methylstyrene and α-isopropylstyrene with ketone 2a gave (S)-α-methylstyrene oxide in 30% ee and α- isopropylstyrene oxide in 58% ee.^3c Several possible spiro and planar transition states for epoxidation with ketone 2 are shown in Figure 3.^2b,3c Spiro transition states (A-D) are generally favored stereoelectronically as a result of the stabilizing interaction of an oxygen lone pair with the π* orbital of the alkene.^1,9,10 However, planar transition states E and G appear to be sterically more favored as compared to spiro transition states. Planar transition states F and H are disfavored both electronically and sterically, thus are unlikely to be significant contributors. Between the two planar transition states E and G, E is likely to be favored over G due to the associative interaction between the phenyl group of the olefin and the oxazolidinone of the ketone catalyst. We hypothesized that planar E might be the major transition state for the epoxidation of α-methylstyrene based on the S configuration of the resulting epoxide obtained with ketone 2a. A higher ee obtained by with α-isopropylstyrene could be due to disfavoring competing spiro D by a larger isopropyl group.^3c Based on these observations, we decided to search for ketone catalysts that can further favor planar E-like transition state to enhance the enantionselectivity for the epoxidation of 1,1-disubstituted terminal olefins. We have found that lactam ketones 3 provide very promising results (Figure 1). Herein we wish to report our studies on this subject.

Figure 2.

Figure 3.

The proposed spiro and planar transition states for the epoxidation of 1,1- disubstituted terminal olefins

Results and Discussion

The synthesis of lactam ketone 3 is outlined in Schemes 1 and 2. Diol 4, prepared from D-glucose as previously reported,¹¹ was treated with BrCH₂COBr to form compound 5, which was then converted to ketone 3a after cyclization and oxidation. Upon introduction of a Boc or Ac group, ketone 3a was converted to ketones 3b and 3c (Scheme 1). Ketones 3d-h were prepared from D-glucose in four steps by Amadori rearrangement,¹² ketalization,^3d formation of the six-membered lactam, and subsequent oxidation (Scheme 2). The X-ray structure of ketone 3d is shown in Figure 4. An overlay of ketones 2b and 3d is shown in Figure 5. In contrast to ketone 2b,^3d the N-phenyl group and the lactam carbonyl group in 3d are not coplanar.

Scheme 1.

Scheme 2.

Figure 4.

The X-ray structure of ketone 3d (stereoview)

Figure 5.

Crystal structure overlay of ketones 2b and 3d (stereoview)

Initial studies on the epoxidation of α-isopropylstyrene with ketone 3d showed that 1,4-dioxane was among the best solvents, giving 94% conversion and 84% ee (Table 1, entry 4). The enantioselectivity was also affected by the N-substituents of ketone catalysts with ketones 3a, 3d, 3e, 3f, and 3h giving the highest enantioselectivity (82-84% ee) (Table 1, entries 7, 4, 10, 11, and 13). Ketone 3d, readily synthesized from inexpensive starting materials, was subsequently investigated for the epoxidation of 1,1-disubstituted terminal olefins. As shown in Table 2, a variety of aryl-substituted 1,1-disubstituted olefins can be effectively epoxidized in good enantioselectivities (62-88% ee). Generally speaking, substrates with bulky alkyl groups at α positions of olefins produce epoxides with higher enantioselectivity than those with small groups. The substituents on the phenyl groups of olefins also have some effects on the enantioselectivities (74-88% ee) (Table 2, entries 7–14). Allylic, homoallylic, and bishomoallylic alcohols are also effective substrates (Table 2, entries 16–21). Up to 88% ee was obtained for 1,1-dialkyl-2-aryl allylic alcohols (Table 2, entries 19–21). A reasonable enantioselectivity (60% ee) was also obtained for a non-aromatic allylic alcohol (Table 2, entry 22).

Table 1.

Asymmetric Epoxidation of α-Isopropylstyrene with Ketones 3^a

entry	ketone	solvent	conv. (%)b	ee (%)b
1	3d	CH3CN/DMM (1/2)	97	71
2		DME	94	81
3		DME/n-BuOH	99	76
4		1,4-dioxane	94	84
5		1,4-dioxane/DME (2/1)	99	80
6		1,4-dioxane/n-BuOH (1/1)	100	78
7	3a		91	82
8	3b		69	71
9	3c		10	nd
10	3e		100	83
11	3f		98	83
12	3g		80	52
13	3h		99	84

^aAll epoxidations were carried out with the olefin (0.2 mmol), ketone 3 (0.06 mmol), Oxone (0.32 mmol), and K₂CO₃ (1.344 mmol) in organic solvent (3 mL) and buffer (0.1 M K₂CO₃/AcOH, pH 9.3; 2 mL) at −10 °C for 2 h.

^bThe conversion and ee were determined by chiral GC (B-DM column).

Table 2.

Asymmetric Epoxidation of 1,1-Disubstituted Terminal Olefins with Ketone 3d^a

entry	substrate	yield (%)b	ee (%)	config.e
1	R = Me	60	62c	(+)−(S)13a
2	R = Et	71	78d	(+)–(S)13a
3	R = n-Pr	90	75d	(+)
4	R = i-Bu	54	74d	(+)
5	R = c-C6H11	62	77c	(+)
6	R = t-Bu	43	86d	(+)
7	X = H	71	84d	(+)
8	X = p-i-Pr	51	82c	(+)
9	X = p-MeO	94	84c	(+)
10	X = p-F	78	74d	(+)
11	X = p-Br	68	78d	(+)
12	X = m-Me	57	82c	(+)
13	X = m-F	74	81d	(+)
14	X = o-F	72	88d	(+)
15		51	66c	(−)–(S)13b
16	n = 1	93	77c	(+)–(R)13c
17	n = 2	47	72c	(+)
18	n = 3	62	74c	(+)
19	R = Me	76	87c	(+)–(S)13a
20	R = Et	85	87d	(+)
21	R,R = (CH2)4	86	88d	(+)–(S)13a
22		78	60d	(+)

^aAll epoxidations were carried out with the olefin (0.2 mmol), ketone 3d (0.06 mmol), Oxone (0.32 mmol), and K₂CO₃ (1.344 mmol) in 1,4-dioxane (3 mL), and buffer (0.1 M K₂CO₃/AcOH, pH 9.3; 2 mL) at −10 °C for 2 h (4 h for entries 6, 11, 13, and 14).

^bIsolated yield except entry 7 which is crude yield.

^cThe ee was determined by chiral HPLC (Chiracel OD column).

^dThe ee was determined by chiral GC (B-DM column).

^eThe absolute configurations were determined by comparing the measured optical rotations and HPLC trace with reported ones.

Ketone 3d gave a similar level of enantioselectivity to ketone 2 for epoxidation of cis-olefins (Table 3, entries 1 and 2), indicating that there still exists an electronic attraction between the amide moiety and the phenyl group of the olefin in spiro transition state I (Figure 6). When 1-phenylcyclohexene was epoxidized with ketone 3d, the (S,S)-epoxide derived from planar L (Figure 7) was obtained with 80% ee while the epoxidation with ketones 2a and 2b gave 43% ee of the (S,S)-epoxide^3c and 25% ee of the (R,R)-epoxide,^3d respectively, suggesting that the six-membered lactam moiety provides a more favorable environment for the attraction between the lactam moiety of the ketone and the phenyl group of the olefin in the planar transition state as compared to ketones 2a and 2b. In the case of 1-phenyl-3,4-dihydronaphthalene, the epoxide resulting from the planar transition state was obtained in as high as 90% ee (Table 3, entry 4), further illustrating the aforementioned attraction in the planar transition state.

Table 3.

Asymmetric Epoxidation of cis- and Trisubstituted Olefins by Ketone 3d^a

entry	substrate	conv. (yield) (%)b	ee (%)	config.g
1		100c(60)	85e	(−)-(1R,2S)3a
2		89d(87)	84f	(+)-(3R,4R)3a
3		99c(89)	80e	(−)-(S,S)2b,3c,d
4		88d(56)	90f	(+)-(1R, 2S)2b

^aAll reactions were carried out with substrate (0.2 mmol), ketone 3d (0.06 mmol for entry 1, 0.04 mmol for entries 2, 3, and 4), Oxone (0.32 mmol), and K₂CO₃ (1.344 mmol) in DME/DMM (3:1, v/v; 3.0 mL) and buffer (0.1 M K₂CO₃-AcOH in 4 × 10⁻⁴ M aqueous EDTA, pH 9.3; 2 mL); For entries 1, 3, and 4, the reaction was carried out at −10 °C for 4 h; For entry 2, the reaction was carried out at 0 °C for 12 h.

^bIsolated yield.

^cThe conversion was determined by GC (B-DM column).

^dThe conversion was determined by ¹H NMR.

^eThe ee was determined by chiral GC (B-DM column).

^fThe ee was determined by chiral HPLC (Chiracel OD column).

^gThe absolute configurations were determined by comparing the measured optical rotations and GC trace with reported ones.

Figure 6.

The proposed competing spiro transition states for the epoxidation of cis-olefins with ketone 3

Figure 7.

The proposed competing spiro and planar transition states for the epoxidation of 1-phenylcyclohexene with ketone 3

The known absolute configurations of selected epoxides (Table 2, entries 1, 2, 15, 16, 19, and 21) are consistent with the notion that the epoxidation proceed mainly via planar transition state P (Figure 8). A bulky R substituent on the olefin disfavors spiro O, thus resulting in higher ee’s as observed. Further improvement of the enantioselectivity will require further disfavoring spiro N and/or planar Q transition states.

Figure 8.

The proposed competing transition states for the epoxidation of 1,1-disubstituted terminal olefins with ketone 3

In summary, a variety of 1,1-disubstituted terminal olefins can be enantioselectively epoxidized using lactam ketone 3d as catalyst and Oxone as oxidant, giving up to 88% ee. Studies indicate that the epoxidation of 1,1-disubstituted terminal olefins with ketone 3 proceeds mainly via a planar transition state. Ketone 3 provides a promising lead for further improvement of the enantioselectivity for this challenging class of olefins.

Experimental

Representive Ketone Synthesis

To a solution of amino alcohol 7d (prepared from D-glucose in two steps)^3d (3.09 g, 10.0 mmol) and Et₃N (1.11 g, 1.54 mL, 11.0 mmol) in dry THF (50 mL), a solution of 2-bromoacetyl bromide (2.22 g, 0.95 mL, 11.0 mmol) in dry THF (10 mL) was added dropwise at rt over 2 h. After the resulting mixture was stirred at rt for 3 h, NaH (95 %, 0.6 g, 23.7 mmol) was added into the reaction mixture carefully. Upon stirring at rt for 0.5 h, the reaction mixture was quenched with MeOH (0.25 mL) and filtered. The filtrate was concentrated and purified by flash chromatography (silica gel, hexane /EtOAc = 1/6) to give lactam 8d as a white solid (1.42 g, 41% yield): mp 198–199 °C; [α]_D²⁵ = −144.6 (c 1.0, CHCl₃); IR (film) 3410, 1661 cm⁻¹; ¹H NMR (400 MHz, CDCl₃) δ 7.21-7.14 (m, 4H), 4.40-4.36 (m, 1H), 4.30-4.21 (m, 4H), 4.12 (d, J = 13.2 Hz, 1H), 3.96 (dd, J = 13.2, 2.8 Hz, 1H), 3.62-3.59 (m, 1H), 3.53-3.48 (m, 1H), 3.10-2.88 (m, 1H), 2.33 (s, 3H), 1.51 (s, 3H), 1.37 (s, 3H); 1³C NMR (100 MHz, CDCl₃) δ 165.6, 138.4, 137.4, 130.1, 125.8, 109.7, 96.2, 76.5, 73.4, 71.7, 62.7, 60.5, 54.2, 28.2, 26.2, 21.2; HRMS Calcd for C₁₈H₂₄O₆N (M+H): 350.1604; Found: 350.1607.

AcOH (0.15 mL) was added to a slurry of lactam 8d (4.8 g, 13.76 mmol), PDC (10.3 g, 27.5 mmol), and 3Å MS (6.5 g) in CH₂Cl₂ (300 mL). Upon stirring at rt for 3 d (no SM left as judged by TLC), the reaction mixture was filtered through a pad of silica gel, and the filter cake was washed with EtOAc. The filtrate was concentrated and purified by flash chromatography (silica gel, hexane/EtOAc = 3/1) to give ketone 3d as a white solid (4.5 g, 95% yield): mp 184–185 °C; [α]_D²⁵ = −86.5 (c 1.0, CHCl₃); IR (film) 1753, 1674 cm⁻¹; ¹H NMR (300 MHz, CDCl₃) δ 7.24-7.18 (m, 4H), 4.86 (d, J = 5.7 Hz, 1H), 4.66-4.64 (m, 1H), 4.49-4.23 (m, 5H), 3.64 (d, J = 13.8 Hz, 1H), 2.36 (s, 3H), 1.47 (s, 3H), 1.43 (s, 3H); ¹³C NMR (75 MHz, CDCl₃) δ 197.7, 165.2, 138.2, 137.7, 130.2, 125.8, 111.0, 96.1, 78.4, 75.7, 63.2, 59.9, 51.9, 27.3. 26.2, 21.3; HRMS Calcd for C₁₈H₂₂NO₆ (M+H): 348.1447; Found: 348.1447; Anal. Calcd. For C₁₈H₂₁NO₆: C, 62.24; H, 6.09. Found: C, 62.02; H, 6.01.

Representative Epoxidation Procedure (Table 2, Entry 19)

To a solution of the olefin (0.324 g, 0.20 mmol), tetrabutylammonium hydrogen sulfate (0.0038 g, 0.010 mmol), and ketone 3d (0.0208 g, 0.06 mmol) in dioxane (3 mL) was added buffer (0.1 M K₂CO₃-AcOH in 4 × 10⁻⁴ M aqueous EDTA, pH = 9.3; 2 mL) with stirring. After the mixture was cooled to −10 °C (bath temperature), a solution of Oxone (0.20 M in 4 × 10⁻⁴ M aqueous EDTA, 1.6 mL) (0.197 g, 0.32 mmol) and a solution of K₂CO₃ (0.84 M in 4 × 10⁻⁴ M aqueous EDTA, 1.6 mL) (0.185 g, 1.344 mmol) were added separately and simultaneously via a syringe pump over a period of 2 h. The reaction mixture was quenched with hexane, extracted with EtOAc, dried over Na₂SO₄, filtered, concentrated, and purified by flash chromatography (silica gel was buffered with 1% Et₃N in organic solvent; hexane/Et₂O = 5/1 as eluent) to give the epoxide as white solid (0.027 g, 76% yield, 87% ee).

Supplementary Material

4_si_004

Supporting Information Available: The synthesis and characterization of ketones 3a-h, the epoxidation procedure and characterization of epoxides, and the X-ray structures of ketones 3c, 3d, and 3g along with the data for the determination of the enantiomeric excess of the epoxides obtained with ketone 3d (126 pages).