Preparation of optically active 4-substituted γ-lactones by lipase-catalyzed optical resolution

Preparation of N-methyl-4- hydroxyalkanamides

We have previously reported the preparation of various N-methyl-4-hydroxyalkanamides from N-methylsuccinimide by the Grignard reaction and subsequent reductive reaction [30]. In this paper, the introduction of various functional groups was investigated (Scheme 1, Table 1). Alkylations with primary alkyl halides yielded the corresponding N-methyl-4-hydroxyalkanamides (rac-1a,e,f) in the yields ranging from 61% to 79% (entries 1, 5, and 6). With aryl bromides, N-methyl-4-hydroxyalkanamides (rac-1c,d,g) were synthesized with yields in the range of 87–94% (entries 3, 4, and 7). By contrast, the purification of N-methyl-4-hydroxyalkanamides (rac-1b,h,i) obtained from secondary and tertiary alkyl bromides was not satisfactory, and the crude yields were very low (entries 2, 8, and 9). These decreases in yield result from the side reactions such as the Wurtz reaction and olefin formation, and the low reactivity is caused by steric hindrance [31, 32]. Accordingly, the reactions with organolithium reagents were attempted (Scheme 2). The reactivity of organolithium reagents is higher than that of Grignard reagents and organozinc reagents [33]. Organolithium reagents can generally be used as nucleophiles. Many reports have been published about effective alkylation using organolithium reagents [34–39]. N-methyl-4-hydroxy-5,5-dimethylhexanamide (rac-1b) substituted with a tert-butyl group was synthesized with 44% yield using tert-butyllithium.

Scheme 1

Synthesis of N-methyl-4-hydroxyalkanamides with Grignard reagent.

Table 1

Preparation of compounds rac-1a–i (see Schemes 1 and 2).

Entry	R-X	Product	Yield (%)	Entry	R-X	Product	Yield (%)
1		rac-1a	68	6		rac-1f	61
2		rac-1b	10a	7		rac-1g	87
3		rac-1c	94	8		rac-1h	6a
4		rac-1d	94	9		rac-1i	10a
5		rac-1e	79

^aCrude product.

Scheme 2

Synthesis of N-methyl-4-hydroxy-5,5-dimethylhexanamide rac-1b.

Lipase-catalyzed enantioselective acetylation of N-methyl-4-hydroxyalkanamides

We have previously reported detailed studies of lipase-catalyzed enantioselective acetylation [30, 40]. The lipase-catalyzed acetylation was performed in diethyl ether using N-alkyl-4-hydroxyalkanamides as substrates, vinyl acetate as acyl donor, and Novozym 435 as lipase. These conditions result in high enantioselectivity, and both enantiomers of various γ-lactones have successfully been obtained with more than 99% enantiopurity. In that previous work, lipase screening has been examined using Novozym 435 (immobilized from Candida antarctica), porcine pancreas lipase (PPL) (from porcine pancreas), lipase from pseudomonas fluorescens, immobilized (LPI) (immobilized from Pseudomonas fluorescens), and Lipozym IM (immobilized from Mucor miehei) in the acetylation of racemic N-methyl-4-hydroxyundecanamide that is similar in structure to rac-1. With a notable exception of Novozym 435, these lipases exhibit no reactivity toward racemic N-methyl-4-hydroxyundecanamide.

To determine the optimal amount of Novozym 435, the reaction was conducted in diethyl ether at room temperature with different ratios of Novozym 435–1.0 mmol of the substrate. It was concluded that 0.4 g is the optimal amount of Novozym 435–1.0 mmol of the substrate. The acetylation with Novozym 435 proceeds well for all substrates except rac-1b, although the time required to reach approximately 50% conversion varies for different starting materials (Scheme 3, Table 2). We have previously studied Novozym 435-catalyzed acetylation of racemic N-methyl-4-hydroxynonanamide [40]. The 50% conversion was reached in 2 h, and (R)- and (S)-γ-nonalactone were obtained with 98% and more than 99% enantiopurities, respectively. In this work, Novozym 435-catalyzed acetylation of rac-1a required 4 h despite the structural small difference between n-pentyl and isopentyl groups. In addition, (R)- and (S)-7-methyl-γ-octalactones (3a and 4a) derived from optically active 1 and 2 were obtained with 96% and more than 99% enantiomeric excesses. The enantiopurity of (R)-7-methyl-γ-octalactone was slightly low compared with that of (R)-γ-nonalactone. These results show that Novozym 435 exhibits not only low substrate affinity but also low substrate selectivity toward rac-1a, which has a sterically bulky R group. Therefore, a long reaction time can be predicted because the substrate possesses a sterically bulky side chain. An attempted acetylation failed for rac-1b substituted with a bulky tert-butyl group (entry 2). It appears that the substrate affinity of Novozym 435 is low for a substrate with a large steric hindrance around the asymmetric carbon atom. The reaction times for N-methyl-4-hydroxy-4-phenylbutanamide (rac-1c, entry 3) and N-methyl-4-hydroxy-4-p-tolylbutanamide (rac-1d, entry 4) are longer than those for rac-1a (entry 1) and N-methyl-4-hydroxy-6-phenylhexanamide (rac-1f, entry 6). By contrast, rac-1c with a phenyl group and rac-1d with a tolyl group are acetylated by Novozym 435. Phenyl and tolyl groups are more bulky than the isopentyl group; thus, a long reaction time is required to reach 50% conversion. In the case of rac-1c and rac-1d, both enantiomers of 4-phenyl-γ-butyrolactone (3c and 4c) and 4-(p-tolyl)-γ-butyrolactone (3d and 4d) were obtained with more than 90% enantiopurities. These optical purities are approximately 5–10% lower than that of rac-1a. These results suggest that the substrate selectivity and the affinity of Novozym 435 toward the high steric hindrance substrate around an asymmetric carbon are low. Naoshima et al. [41] explained the enantioselectivity of lipase by using computer modeling. They measured the C-O distance between the carbonyl carbon atom of the acetyl group in the substrate and the oxygen atom at the active center of lipase. The large difference of the C-O distance among each enantiomer correlated with high enantioselectivity. It appears that for the large steric hindrance around asymmetric carbon atom, the hydroxy group in the substrate and the active center in Novozym 435 are difficult to be approached. It can be suggested that the enantioselectivity of Novozym 435 toward the substrate with large steric hindrance decreases. Although approximately 50% conversion was reached at 11 h for N-methyl-4-hydroxy-4-(p-anisyl)butanamide (rac-1g) with a p-anisyl group, both 4-(p-anisyl)-γ-butyrolactones (3g and 4g) exhibit racemic similarities (entry 7). Phenyl, tolyl, and p-anisyl groups are structurally similar. However, only the reaction of rac-1g gave racemic γ-lactones (3g and 4g), and this result can be attributed to the presence of the methoxy group. The reaction times required for rac-1c and rac-1d to reach 50% conversion by Novozym 435-catalyzed acetylation were 12 and 20 h, respectively, and that for rac-1g took 11 h. These substrates require comparably long reaction time to reach 50% conversion compared with rac-1a and rac-1f. As mentioned earlier, it was assumed that Novozym 435 shows higher substrate affinity toward the small R group in substrates such as rac-1a and 1f. The reaction time of rac-1g was similar to rac-1c. These results also show that Novozym 435 exhibits low affinity toward the bulky substrates around the asymmetric carbon atom. By contrast, it can be suggested that the high substrate affinity of Novozym 435 is due to hydrogen bonding between the oxygen at the methoxy group and the amino acid residues that constitute the lipase. Both enantiomers of rac-1g can be incorporated into the active site by this hydrogen bonding, which causes the observed lack of enantioselectivity. Novozym 435-catalyzed acetylation of all rac-1 except rac-1g progressed with more than 90% enantioselectivity. The stereoselectivity of Novozym 435 varies with structural differences of the R group. The absolute configurations of γ-lactones prepared from hydroxyamide 1 are (R)-form for rac-1a with an isopentyl group and rac-1f with a phenethyl group. By contrast, the corresponding γ-lactones derived from 1 have (S)-configuration in the case of rac-1c with a phenyl group, rac-1d with a tolyl group, and rac-1e with a benzyl group. As shown in Scheme 3, the hydroxyl group is acetylated by Novozym 435 in all substrates. Previously, we have reported a synthetic methodology of the chiral γ-lactones, which combines Novozym 435-catalyzed reaction and Mitsunobu reaction [42, 43]. In this work, Novozym 435-catalyzed stereoselective hydrolysis of N-benzyl-4-acetoxyalkanamides was conducted. The hydrolysis was conducted at 60°C in diisopropyl ether; the reaction progressed with more than 90% enantioselectivity. However, the 50% conversion was reached after a relatively long period of 24–30 h. In summary, Novozym 435 shows high enantioselectivity for both acetylation and hydrolysis.

Scheme 3

Novozym 435-catalyzed stereoselective acetylation of rac-1a–g.

Lipase-catalyzed enantioselective hydrolysis of N-methyl-4-acetoxyalkanamides

In the acetylation using Novozym 435, rac-1b was inert (Table 2, entry 2). Enantioselectivity was not observed for rac-1g, although approximately 50% conversion was reached in 11 h (Table 2, entry 7). Novozym 435-catalyzed hydrolysis of rac-2 was investigated to prepare optically active lactones 3b, g and 4b, g (Scheme 4, Table 3). In all cases, the time required for the hydrolysis to reach approximately 50% conversion is much longer than that for acetylation. Novozym 435 is inert toward the hydrolysis of rac-2b as well as the acetylation of rac-1b (entry 2). It appears that the bulky tert-butyl group is not compatible with the substrate specificity of Novozym 435. Although rac-2g is hydrolyzed, the obtained lactone 3g/4g is racemic (entry 7). Because rac-2c and rac-2d are hydrolyzed enantioselectively, it can be suggested that the anisyl group of rac-2g is a factor that does not promote enantioselectivity. The reaction mechanism of lipase was reported [44, 45].

Scheme 4

Novozym 435-catalyzed stereoselective hydrolysis of rac-2a–g.

Preparation of racemic N-methyl-4-hydroxyalkanamides rac-1a,c–g

Organomagnesium reagents were freshly prepared by the slow addition of the corresponding bromides or chlorides (16.5 mmol) in Tetrahydrofuran (THF) (50 mL) onto magnesium turnings (0.37 g, 15.0 mmol) previously activated with a crystal of iodine. Under vigorous stirring and cooling with an ice bath, N-methylsuccinimide (1.13 g, 10.0 mmol) dissolved in THF (30 mL) was then added to the suspension, and the mixture was stirred for 8 h at room temperature. The residual Grignard reagent was hydrolyzed by gradual addition of ice (10 g) and saturated aqueous NH₄Cl (100 mL). The organic phase was separated, and the aqueous phase was extracted with CHCl₃ (4×50 mL). The combined extracts were washed with water and dried with MgSO₄. The solvent was evaporated under reduced pressure, and the residue was not purified. NaBH₄ (0.76 g, 20 mmol) in methanol (20 mL) was added to the crude product with stirring, and the mixture was stirred for 1 h at room temperature. The solvent was evaporated under reduced pressure, and water (50 mL) was added. The aqueous phase was extracted with CHCl₃ (4×50 mL). The combined extracts were washed with water and dried with MgSO₄. The solvent was evaporated, and the residue was purified by flash chromatography on silica eluting with EtOAc to give the corresponding N-methyl-4-hydroxyalkanamide 1.

Preparation of N-methyl-4-hydroxy- 5,5-dimethylhexanamide (1b)

A solution of tert-butyllithium in pentane (6.29 mL, 10.0 mmol) was added dropwise to a solution of N-methylsuccinimide (1.13 g, 10.0 mmol) in THF (20 mL) at -78°C under argon atmosphere, and the mixture was stirred at the same temperature for 2 h. The reaction mixture was poured onto saturated NH₄Cl at 0°C and extracted with EtOAc (5×50 mL). The combined organic layers were washed with water, dried with MgSO₄, and concentrated in vacuo. The crude product was not purified. NaBH₄ (0.76 g, 20.0 mmol) was added to a solution of the crude product in MeOH (20 mL) at 0°C, and the mixture was stirred for 1 h. The solvent was evaporated, and water (50 mL) was added to the residue. The aqueous phase was extracted with EtOAc (5×50 mL), and the combined organic layers were washed with water, dried with MgSO₄, and concentrated in vacuo. The crude product was purified by crystallization from n-hexane to give N-methyl-4-hydroxy-5,5-dimethylhexanamide (1b, 0.76 g, 44%) as a colorless solid; mp 117–118°C (dec); R_f=0.30 (eluent: CHCl₃-MeOH, 10:1, v/v); IR: 3292 (N-H, O-H), 2954 (-CH₃), 2932 (-CH₂-), 2868 (-CH₃), 2846 (-CH₂-), 1645 cm^-1 (-NHC=O); ¹H NMR: δ 0.91 (s, 9H, -C(CH₃)₃), 1.60 (m, 1H, -NHC(=O)CH₂CH₂-), 1.89 (m, 1H, -NHC(=O)CH₂CH₂-), 2.37 (m, 2H, -NHC(=O)CH₂-), 2.63 (br, 1H, -OH), 2.82 (d, 3H, J = 4.6 Hz, -NHCH₃), 3.20 (m, 1H, -CH(OH)-), 5.66 (br, 1H, -NH-); ¹³C NMR: δ 25.6 (-C(CH₃)₃), 26.4 (-NHCH₃), 27.0 (-CH₂-), 34.1 (-C(=O)CH₂-), 35.0 (-C(CH₃)₃), 79.7 (-CHOH), 174.6 (-NHC(=O)-). HRMS (ESI). Calcd for C₉H₂₀NO₂ (M+H)⁺: m/z 174.1494. Found: m/z 174.1513.

General procedure for Novozym 435-catalyzed acetylation and lactonization

Novozym 435 (0.4 g) was added to a racemic mixture of N-methyl-4-hydroxyalkanamides (1, 1.0 mmol) and vinyl acetate (2.0 mmol) in diethyl ether (20 mL). After stirring at room temperature for a period specified in Table 2, the reaction mixture was filtered. Concentration under a reduced pressure followed by flash chromatography of the residue on silica gel (eluent: EtOAc) afforded the corresponding optically active N-methyl-4-hydroxyalkanamide 1a–g and N-methyl-4-acetoxyalkanamide 2a–g. Then NaOH (2.0 g) was added to a solution of 1a–g or 2a–g in methanol (20 mL), and the mixture was heated under reflux for 3 h. After cooling, methanol was removed under reduced pressure and water (50 mL) was added. The aqueous phase was acidified with 10% HCl to pH 3.0, and the mixture was stirred for 8 h then extracted with EtOAc (4×20 mL). The combined organic layers were washed with water, dried with MgSO₄, and concentrated under reduced pressure. The residue was purified by flash chromatography eluting with n-hexane-EtOAc, 4:1, to give the corresponding γ-lactone (3 or 4) as a colorless oil. The enantiomeric excess was measured by using the chiral GC analysis. The absolute configurations of 3 and 4 were determined by optical activity compared with the literature data.

General procedure for Novozym 435-catalyzed hydrolysis

Racemic N-methyl-4-acetoxyalkanamides rac-2a–g were prepared almost quantitatively from N-methyl-4-hydroxyalkanamides rac-1a–g by acetylation using acetic anhydride [30]. Briefly, a mixture of racemic N-methy-5-acetoxyalkanamide (rac-2a–g, 1.0 mmol), methanol (3.0 mmol, 0.10 g), Novozym 435 (0.4 g), and diethyl ether (20 mL) was stirred at 40°C for a period specified in Table 3, then filtered to remove Novozym 435, and concentrated. The purification of the crude product by silica gel column chromatography eluenting with EtOAc gave optically active N-methyl-4-hydroxyalkanamide 1a–g and N-methyl-4-acetoxyalkanamide 2a–g. Lactonization is described earlier.

Determination of enantiomeric excess

Enantiomeric excesses of optically active γ-lactone 3 and 4 were measured by chiral GC. General GC conditions: chiral column, CycloSil B; injector temperature, 250°C; detector temperature, 250°C; He gas, 2.0 mL/min.

7-Methyl-γ-octalactone 3a and 4a: Oven temperature, 140°C (isothermal); retention time, 8.5 min for (R)-3a, 8.8 min for (S)-4a.

4-Phenyl-γ-butyrolactone 3c and 4c: Oven temperature, 150°C (isothermal); retention time, 17.2 min for (S)-3c, 19.0 min for (R)-4c.

4-p-Tolyl-γ-butyrolactone 3d and 4d: Oven temperature, 150°C (isothermal); retention time, 27.0 min for (S)-3d, 30.4 min for (R)-4d.

5-Phenyl-γ-pentalactone 3e and 4e: Oven temperature, 140°C (isothermal); retention time, 27.3 min for (S)-3e, 28.2 min for (R)-4e.

6-Phenyl-γ-hexalactone 3f and 4f: Oven temperature, 140°C (isothermal); retention time, 76.0 min for (R)-3f, 77.8 min for (S)-4f.

4-p-Anisyl-γ-butyrolactone 3g and 4g: Oven temperature, 160°C (isothermal); retention time, 39.0 min for 3f, 42.6 min for 4f.

Specific rotation of optically active amides 1 and 2

The absolute configuration and the enantiomeric excesses were determined by using the values of the corresponding lactones 3 and 4.

(R)-N-methyl-4-hydroxy-7-methyloctanamide [(R)-1a]: [α]²⁵_D=-4.8 (c 1.0, MeOH, 96% e.e.).

(S)-N-methyl-4-hydroxy-4-phenylbutanamide [(S)-1c]: [α]²⁵_D=-40.7 (c 1.0, CHCl₃, 91% e.e.).

(S)-N-methyl-4-hydroxy-4-(p-tolyl)butanamide [(S)-1d]: [α]²⁵_D= -48.6 (c 1.0, CHCl₃, 95% e.e.).

(S)-N-methyl-4-hydroxy-5-phenylpentanamide [(S)-1e]: [α]²⁵_D= +4.7 (c 1.0, CHCl₃, >99% e.e.).

(R)-N-methyl-4-hydroxy-6-phenylhexanamide [(R)-1f]: [α]²⁵_D=+8.4 (c 1.0, MeOH, >99% e.e.).

(S)-N-methyl-4-acetoxy-7-methyloctanamide [(S)-2a]: [α]²⁵_D=+8.0 (c 1.0, MeOH, >99% e.e.).

(R)-N-methyl-4-acetoxy-4-phenylbutanamide [(R)-2c]: [α]²⁵_D=+58.8 (c 1.0, CHCl₃, 94% e.e.).

(R)-N-methyl-4-acetoxy-4-(p-tolyl)butanamide [(R)-2d]: [α]²⁵_D=+65.7 (c 1.0, CHCl₃, 93% e.e.).

(R)-N-methyl-4-acetoxy-5-phenylpentanamide [(R)-2e]: [α]²⁵_D=+13.1 (c 1.0, CHCl₃, 95% e.e.).

(S)-N-methyl-4-acetoxy-6-phenylhexanamide [(S)-2f]: [α]²⁵_D=+8.1 (c 1.0, MeOH, 99% e.e.).

Assignment of absolute configuration for lactones 3 and 4

The absolute configuration of γ-lactones 3 and 4 was determined by comparison of the sign of the measured specific rotation with that in the literature.

(S)-7-Methyl-γ-octalactone 4a: [α]²⁵_D=-46.7 (c 1.0, MeOH, >99% e.e); Lit. [α]²⁰_D=-38.6° (c 0.26, MeOH) [46].

(R)-4-Phenyl-γ-butyrolactone 4c: [α]²⁵_D=+32.8 (c 1.0, CHCl₃, 94% e.e.); Lit. [α]²⁵_D=+20.3° (62% e.e.) [47].

(R)-4-p-Tolyl-γ-butyrolactone 4d: [α]²⁵_D=+25.7 (c 1.0, CHCl₃, 93% e.e.); Lit. [α]²⁵_D=+10.8° (91% e.e.) [48].

(S)-5-Phenyl-γ-pentalactone 3e: [α]²⁵_D=+19.0 (c 1.0, CHCl₃, >99% e.e.); Lit. [α]²⁵_D=+24.7° (c 1, CHCl₃, 97% e.e.) [29].

(R)-6-Phenyl-γ-hexalactone 3f: [α]²⁵_D=+65.8 (c 1.0, MeOH, >99% e.e.); Lit. [α]²⁵_D=+39.2° (c 0.2–1.0, MeOH, >99% e.e.) [46].