2025 Volume 73 Issue 6 Pages 520-525
This study investigated the application of Corey–Bakshi–Shibata (CBS) catalysts to asymmetric reduction of 2′-aminoacetophenone derivatives and determined their absolute configuration using the modified Mosher and microcrystal electron diffraction methods. The results reveal that stereoselective CBS reduction is effective on 2′-amino-3′-methylacetophenone, yielding secondary alcohol, and that the reaction proceeds stereoselectively, even on the gram scale. Moreover, (R)-(–)-secondary alcohol configuration was obtained using an (S)-Me-CBS catalyst, and the stereoselectivity of the reduction followed a previously proposed reaction mechanism for acetophenone derivatives. Thus, this study demonstrated that secondary alcohol could be obtained with the expected stereoselectivity, although requiring a slightly higher amount of the CBS catalyst. The study findings suggest that the amino group of aniline may affect the progress of CBS reduction but does not significantly affect the transition state.
Asymmetric synthesis of secondary alcohols has been of significant research interest,1–6) and the asymmetric reduction of prochiral ketones, one of the simplest chemical transformations, efficiently affords optically active secondary alcohols. In the presence of an appropriate chiral catalyst, hydride is selectively added to the carbonyl moiety from one of the two enantiofaces. Corey–Bakshi–Shibata (CBS) catalyst is one of the most useful catalysts, applicable to a wide range of substrates.1) The enantioselective CBS reduction is a prominent method for the asymmetric synthesis of chiral medicines, catalytic ligands, and complex natural products.7–14) As the transition state of the CBS reduction reaction has been elucidated, the stereochemistry of the resulting secondary alcohol can be predicted based on the stereochemistry of the CBS catalyst.
During our research, the stereoselective reduction of acetyl groups substituted at the ortho-position of aniline derivative 1 needed to be investigated. To the best of our knowledge, no reports on the application of CBS catalyst to the reduction of acetophenones bearing unprotected amino groups exist. Therefore, this study investigated the applicability of the CBS reduction to unprotected aniline, 2′-amino-3′-methylacetophenone 1 (Chart 1). Furthermore, the absolute configuration of secondary alcohol product 2 was determined based on the modified Mosher and microcrystal electron diffraction (microED) methods. Our results reveal that the aniline nitrogen of 1 may not be involved in controlling the stereochemistry of the CBS reduction transition state.
This study examined CBS reduction on the basis of a study by Zhang et al.15) and observed that the reduction of 2′-amino-3′-methylacetophenone 1, which was derived from 2-amino-3-methylbenzoic acid 3, with (S)-5,5-diphenyl-2-methyl-3,4-propano-1,3,2-oxazaborolidine ((S)-Me-CBS catalyst) proceeded with high enantioselectivity. Table 1 summarizes the results of the study. The reduction of 1 with 0.7 equivalent (equiv.) of BH3·tetrahydrofuran (THF) in the presence of 0.2 equiv. of (S)-Me-CBS catalyst in THF at 0°C for 2 h resulted in the corresponding secondary alcohol 2 in 59% yield with 61% enantiomeric excess (ee) (Table 1, entry 1). After surveying the reaction conditions, the use of 0.7 equiv. of catalyst was observed to produce the best enantioselectivity (Table 1, entry 3). Furthermore, the optimum temperature was −10°C (Table 1, entry 5). However, the yield did not improve even with 0.7 equiv. amount of the catalyst; this may be due to certain interactions between the unprotected amino group of aniline and the catalyst. Furthermore, extending the reaction time had no effect, and therefore no further improvement in yield was expected; despite this study limitation, further investigations were conducted to achieve satisfactory enantioselectivity on the gram scale (Table 1, entries 7, 8). Subsequently, 33 mmol (4.9 g) of 1 was subjected to CBS-reduction to provide 2 in 46% yield (2.3 g) and 87% ee. To improve the optical purity, recrystallization of 2 from dichloromethane was investigated, resulting in an amorphous compound 2 in a yield of 34% with 99.7% ee and an [α]D value of –15.8 (c 0.21, CHCl3).
![]() |
a) Reaction was performed using 1 mmol of 1. b) Reaction was performed using 10 mmol of 1. c) Reaction was performed using 33 mmol of 1 for 17 h.
Subsequent to obtaining optically pure (−)-2 by using (S)-Me-CBS catalyst, this study examined whether the stereoselective reduction proceeded according to the mechanism proposed by Corey et al.1) As this study is the first to realize the synthesis of optically pure 2, its absolute configuration has not yet been determined. However, X-ray crystal structure analysis was not possible here because (−)-2 did not form a single crystal. Therefore, to determine the absolute configuration of (−)-2, the modified Mosher’s method,16) which is widely accepted and has been applied to determine the absolute configurations of some natural products,16–22) was employed. By employing this method, the absolute configuration of organic compounds containing a secondary alcohol moiety can be revealed on the basis of the 1H-NMR spectra of their methoxy(trifluoromethyl)phenylacetic acid (MTPA) esters. Prior to MTPA esterification, the amino group of (rac)-2, which was prepared via the NaBH4 reduction of 1, was selectively protected with a Boc group to provide amide 4 quantitatively (Chart 2).
To determine the configuration of secondary alcohol of 4, (rac)-4 was reacted with (S)-MTPA-Cl to obtain (S,R)/(R,R)- MTPA esters 5 as diatereomixtures in 32% yield (5a:5b = 1 : 2). Because of the difference in priorities, the stereochemistry of the ester formed by the reaction of (S)-MTPA-Cl changed to (R). Diastereomers 5 were separated using HPLC to impart the faster eluting compound as 5a and the slower eluting compound as 5b. The modified Mosher’s method is based on a Mosher’s concept that MTPA ester groups exist in a conformation in which the proton at an asymmetric center and the C=O carbonyl bond and the CF3 group are located in the same plane. In the (R,R)-MTPA ester (A in Chart 2), Me (in blue) was shielded by the benzene ring of the MTPA ester and shifted upfield. Similarly, in the (S,R)-MTPA ester (B in Chart 2), the protons of the Ar group (in red) are shielded by the benzene ring of the MTPA ester and shift upfield. In contrast, the protons of the Ar group (in red) in (R,R)-MTPA (A in Chart 2) and Me (blue) in the (S,R)-MTPA ester (B in Chart 2) were not upfield. Therefore, Δδ (Δδ = δ(R,R) – δ(S,R)) for the Me group should have negative values (Δδ < 0), and Δδ (Δδ = δ(R,R) – δ(S,R)) for protons of the Ar group should have positive values (Δδ > 0). The configurational properties of 5a and 5b were characterized using 1H-NMR spectroscopy in methanol-d4 based on the modified Mosher’s method. The observed chemical shift and Δδ (5b–5a) values are presented in Table 2.
![]() |
The results presented in Table 2 indicate that the absolute configurations of 5a/5b are (S,R)/(R,R). Subsequently, 5a and 5b were hydrolyzed to the corresponding (S)−2 and (R)−2 in good yields by dissolving in 2n NaOH and methanol and stirring for 3 h at 40°C (Chart 3). Furthermore, the optical rotations of (S)−2 and (R)−2 were determined to be (+) and (−), respectively. Therefore, the absolute configuration and optical rotation of each enantiomer of 2 were determined.
However, when a tert-butoxycarbonyl (t-Boc) group is near the ester moiety, as in 5, causing significant steric hindrance, the conformation of the ester may deviate from the Mosher’s concept. Additionally, the limited number of protons for which chemical shift values could be compared between 5a and 5b is another concern. However, 2 appeared amorphous and did not crystallize. Therefore, the recently developed microcrystal electron diffraction (microED) method was employed.
Determination of Absolute Configuration of Aniline Derivative 2 Using the MicroED MethodMicroED is less practical for chirality analysis because of the short wavelength and high-energy properties of electrons compared with X-rays. Palatinus et al. proposed a dynamic diffraction theory for the structural analysis of continuous rotation three-dimensional (3D) electron diffraction data.23) By considering multiple scattering events during microED data processing and refinement, they determined the crystal structure and absolute stereochemistry of a pharmaceutical cocrystal of sofosbuvir and l-proline. This new approach elucidates the absolute stereochemistry of small molecules.23–26) Therefore, this study used the microED method to elucidate the absolute configuration of (−)-2. Subsequent to the refinement of the R-enantiomer, an inverted model was created, and without changing any parameters, it was refined using a dynamic refinement approach. The correct enantiomer was directly determined by comparing the R-values of the two refinement processes. The enantiomer with lower R-values is designated the correct configuration, while the higher R-values correspond to the incorrect enantiomer. In this study, as the R-values of the R-enantiomer model were lower than those of the S-enantiomer, the absolute configuration of (−)-2 was determined as (R)27) (Supplementary Table S1). The absolute structure of (−)-2, determined to be (R) (Fig. 1), was consistent with the absolute configuration determined using the modified Mosher’s method.
Considering that (R)-2 was obtained via stereoselective reduction using the (S)-Me-CBS catalyst, the reaction mechanism was investigated further (Chart 4). According to the general mechanistic model developed for the reduction of the acetyl moiety of acetophenone derivatives,28) the initial step in the pathway is the coordination of BH3 to the Lewis basic nitrogen atom on the α face of the CBS catalyst to form cis-fused oxazaborolidine⋅BH3 complex A. The strongly Lewis acidic complex A should bind to 1 in forms of I and II, with form I preferred as it minimizes unfavorable steric interactions between oxazaborolidine and 1. Consequently, form I aligns the electronically deficient carbonyl carbon atom and coordinates BH3 for a stereoelectronically favorable, face-selective hydride transfer via a six-membered transition state. Finally, hydrolysis under acidic conditions afforded (R)-2 stereoselectivity.
As described above, the stereoselectivity observed in this study can be interpreted based on the previously proposed considerations of stereoselectivity in acetophenone derivatives. This suggests that the amino group in 1 has little effect on the process that determines the stereoselectivity of the reduction. However, the moderate yields and the requirement of 0.7 equiv. of the CBS catalyst imply that the amino group in 1 affects the catalytic cycle.
Many compounds, including pharmaceuticals, agricultural chemicals, and industrial raw materials, contain amino groups, and the asymmetric hydrogenation of ketones is an important process for the production of optically active functional organic compounds on a small to an industrial scale. This study demonstrated that stereoselective CBS reduction is applicable to the unprotected aniline 2′-amino-3′-methylacetophenone 1. The CBS reduction reaction proceeded on the gram scale, and the absolute configuration of (+)/(−)-1-(2-amino-3-methylphenyl) ethanol 2 was determined to be S/R using the modified Mosher and microED methods. The absolute stereochemistry of the resulting secondary alcohol is consistent with the reaction mechanism proposed for acetophenone derivatives. However, the moderate yields obtained using higher amounts of catalyst suggest that the amino group of aniline 1 may affect the progress of the reaction but does not significantly affect the transition state. The study findings reveal that, as the stereoselective CBS reduction proceeds with general stereoselectivity for unprotected aniline 1, it is feasible also for similar amino-unprotected compounds. Future studies will report on the asymmetric synthesis of specific bioactive compounds using (R)-2 as the key starting material.
Materials used in the experiments were produced from commercial suppliers. NMR spectra were recorded on a spectrometer (JEOL Ltd., Tokyo, Japan and Bruker, Billerica, MA, U.S.A.) operating at 600 MHz for 1H-NMR and 150 MHz for 13C-NMR. Chemical shifts were expressed in ppm relative to tetramethylsilane as an internal standard, and the coupling constants (J) were reported in hertz (Hz). Abbreviated representations of the splitting patterns are as follows: singlet (s), doublet (d), triplet (t), quartet (q), quintet (quin), multiplet (m), and broad (br). IR spectra were acquired using a Fourier-transform IR spectrometer equipped with an attenuated total reflectance accessory (ATR; diamond). High-resolution mass spectra were obtained in an electrospray ionization mode. Melting points were recorded using a melting point apparatus and retained. Optical rotation was determined using a digital polarimeter. Analytical thin-layer and column chromatography was performed on precoated, glass-backed silica gel plates (Merck silica gel 60 F254: Merck KGaA, Darmstadt, Germany) and silica gel (60 µm), respectively. The extracted solutions were dried over anhydrous Na2SO4. Solvents were evaporated under reduced pressure.
Preparation of 2′-Amino-3′-methylacetophenone 1For preparing 2′-amino-3′-methylacetophenone 1, commercially available 2-amino-3-methylbenzoic acid 3 was methylated using methyllithium in moderate yields (Chart 5).
Methyllithium (3.1 M in diethoxymethane, 38.6 mL, 120 mmol) was added to a stirred solution of 2-amino-3-methylbenzoic acid (5.17 g, 34.2 mmol) in cyclopentyl methyl ether (47 mL) at 0°C. The mixture was stirred at 0–25°C for 12 h. Subsequently, the mixture was treated with saturated NH4Cl, and compounds were extracted using dichloromethane, washed with brine, dried over Na2SO4, and concentrated. The crude product was purified via column chromatography (ethyl acetate/hexane = 1 : 9) to obtain 1 as a yellow solid with a yield of 3.33 g (65%). The spectral data are consistent with the results reported in the literature.29)
Synthesis of 2-Amino-α,3-Dimethylbenzenemethanol (2)30)Sodium borohydride (242 mg, 6.23 mmol) was added to a stirred solution of 1 (465 mg, 3.12 mmol) in methanol (47 mL) at 25°C. The mixture was stirred for 25 min, and then treated with 2 n NaOH and concentrated. The concentrate was extracted using dichloromethane, washed with brine, dried over Na2SO4, and concentrated. The crude product was purified via column chromatography (ethyl acetate/hexane = 4 : 6) to afford 2 as a white solid with a yield of 446.0 mg (95%). The spectral data are consistent with the results reported in the literature.30)
Chiral HPLC separation: less polar:
A dried 300-mL round-bottom flask was charged with borane–tetrahydrofuran complex (1.0 M in THF, 23 mL, 23 mmol) and THF (109 mL), and the solution was cooled to −10°C. (S)-l-methyl-3,3-diphenyltetrahydro-1H,3H-pyrrolo [1,2-c]1,3,2)oxazaborole (6.3 g, 23 mmol, 0.70 equiv.) was then added to the reaction flask, and the mixture was stirred for 13 h at −10°C. Compound 1 (4.9 g, 33 mmol) in THF was added in five steps over a period of 1.5 h. The mixture was stirred for 25 min and then quenched with methanol and water. Finally, the compound was extracted using ethyl acetate. The organic layer was washed with water and brine, dried over Na2SO4, and concentrated. The concentrate was purified using silica gel column chromatography (ethyl acetate/dichloromethane = 1 : 3), and to improve optical purity, the resulting compound was recrystallized twice at 0°C using dichloromethane, and compound (R)-2 was collected as a white solid with a yield of 772.6 mg (16%, 99% ee).
N,N-Diisopropylethylamine (417 µL, 2.45 mmol) and di-tert-butyl dicarbonate (535 mg, 2.45 mmol) were added to a stirred solution of 2 (309 mg, 2.04 mmol) in tetrahydrofuran (20.4 mL) at 25°C. The mixture was then stirred under reflux for 15 h. After cooling the mixture to 25°C, di-tert-butyl dicarbonate (535 mg, 2.45 mmol) was added and stirred under reflux for 6 h. Subsequently, the mixture was treated with water and concentrated. The concentrate was extracted using dichloromethane, washed with brine, dried over Na2SO4, and concentrated. The crude product was purified via column chromatography (ethyl acetate/hexane = 3 : 2) to afford 4 as a colorless oil with a yield of 512.5 mg (quantitative).
IR (ATR): 1698 cm−1 (CO).
1H-NMR (600 MHz, CDCl3) δ = 7.29 (d, J = 7.2 Hz, 1H), 7.19–7.15 (m, 2H), 6.44 (s, 1H), 5.09–4.98 (m, 1H), 2.87 (s, 1H), 2.27 (s, 3H), 1.51 (brs, 9H), 1.49 (d, J = 6.6 Hz, 3H).
13C-NMR (150 MHz, CDCl3) δ = 154.7, 141.2, 135.6, 133.2, 130.0, 127.1, 124.0, 80.4, 66.6, 28.3, 22.3, 18.4.
HRMS: m/z [M + Na]+ Calcd for C14H21NNaO3: 274.1414; Found: 274.1414.
Synthesis of 1-(2-((tert-Butoxycarbonyl)amino)-3-methylphenyl)ethyl-(R)-3,3,3-trifluoro-2-methoxy-2-phenylpropanoate (5)4-Dimethylaminopyridine (346.1 mg, 2.83 mmol) and (S)-(+)-α-methoxy-α-(trifluoromethyl)phenyl-acetyl chloride (424 µL, 2.27 mmol) were added to a stirred solution of 4 (474.6 mg, 1.89 mmol) in N,N-dimethylformamide (19 mL) at 0°C. The mixture was stirred at room temperature for 17 h and then treated with water. The compounds were extracted using diethyl ether, was washed with brine, dried over Na2SO4, and concentrated. The crude product was purified via column chromatography (ethyl acetate/hexane = 1 : 3) to afford a mixture 5a and 5b as a colorless oil with a yield of 283.5 mg (32%). The mixture of 5a and 5b was separated using CHIRALPAK® IE (ethanol/hexane = 5 : 95).
(S)-1-(2-((tert-Butoxycarbonyl)amino)-3-methylphenyl)-ethyl-(R)-3,3,3-trifluoro-2-methoxy-2-phenylpropanoate (5a)Colorless oil.
IR (ATR): 1718 cm−1 (CO).
1H-NMR (600 MHz, CDCl3) δ = 7.39–7.35 (m, 1H), 7.33–7.29 (m, 4H), 7.19 (d, J = 7.8 Hz, 1H), 7.10 (t, J = 7.8 Hz, 1H), 7.03 (d, J = 7.8 Hz, 1H), 6.34 (s, 1H), 6.27–6.18 (m, 1H), 3.54 (s, 3H), 2.27 (s, 3H), 1.63 (d, J = 6.0 Hz, 3H), 1.52 (s, 9H).
13C-NMR (150 MHz, CDCl3) δ = 165.9, 154.1, 137.2, 136.9, 133.0, 132.1, 130.8, 129.5, 128.3, 127.4, 127.2, 124.4, 123.3 (q, JC-F = 286.8 Hz), 84.4 (q, JC-F = 27.5 Hz), 80.0, 72.0, 55.5, 28.3, 20.9, 18.3.
HRMS: m/z [M + Na]+ Calcd for C24H28F3NNaO5: 490.1812; Found: 490.1812.
(R)-1-(2-((tert-Butoxycarbonyl)amino)-3-methylphenyl)-ethyl-(R)-3,3,3-trifluoro-2-methoxy-2-phenylpropanoat (5b)IR (ATR): 1698 cm−1 (CO).
1H-NMR (600 MHz, CDCl3) δ = 7.40–7.37 (m, 3H), 7.35–7.32 (m, 2H), 7.24–7.17 (m, 3H), 6.44 (s, 1H), 6.30–6.21 (m, 1H), 3.44 (s, 3H), 2.28 (s, 3H), 1.57 (d, J = 7.8 Hz, 3H), 1.51 (s, 9H).
13C-NMR (150 MHz, CDCl3) δ = 166.2, 154.0, 137.1, 137.0, 133.1, 132.2, 130.9, 129.6, 128.4, 127.4, 124.6, 123.3 (q, JC-F = 286.8 Hz), 84.4 (q, JC-F = 27.5 Hz), 80.1, 72.0, 55.3, 28.3, 20.9, 18.3; several signals overlap.
HRMS: m/z [M + Na]+ Calcd for C24H28F3NNaO5: 490.1812; Found: 490.1812.
Synthesis of (S)-2-Amino-α,3-dimethylbenzenemethanol ((S)-2a)30)2 n NaOH (0.5 mL) was added to a stirred solution of 5a (5.0 mg, 0.0011 mmol) in MeOH (0.5 mL) at room temperature. The mixture was stirred at 40°C for 3 h. Subsequently, the mixture was concentrated. The concentrate was extracted using ethyl acetate. The organic layer was washed with 2 n NaOH and brine, dried over Na2SO4, and concentrated. The crude product was purified via preparative thin-layer chromatography (diethyl ether/hexane = 9 : 1) to afford 2a as a white solid with a yield of 1.5 mg (92%). The spectral data are consistent with the results reported in the literature.30)
Compound 2b was prepared according to a similar procedure described for the preparation of 2a from 5a and exhibited a yield of 1.5 mg (93%). The spectral data are consistent with the results reported in the literature.30)
This work was partly supported by a Grant-in-Aid for Scientific Research (C) (22K06537) from the Japan Society for the Promotion of Science. This research was supported in part by AMED under Grant No. A476ATR. H.A. give thanks for the support of JST SPRING, Grant No. JPMJSP2151. The authors acknowledge ReadCrystal Technology Co. for their prominent works in MicroED data collection and analysis.
The authors declare no conflict of interest.
This article contains supplementary materials.