Stereoselective Synthesis of Highly Functionalized 2,3-Dihydro-4-pyranones Using Phosphine Oxide as Catalyst

Abstract

2,3-Dihydro-4-pyranones were synthesized stereoselectively using a chiral phosphine oxide as the catalyst. The phosphine oxide sequentially activated silicon tetrachloride and promoted the double aldol reaction of 4-methoxy-3-buten-2-one with aldehydes. Subsequent stereoselective cyclization afforded the corresponding highly functionalized 2,3-dihydro-4-pyranones bearing three contiguous chiral centers in good yields and with high diastereo- and enantioselectivities.

Pyran and its derivatives are common and versatile structures in bioactive natural products and medicines.¹⁾ Therefore, it is important to develop efficient methods for the synthesis of diverse pyran structures.^2–5) Recently, we reported the synthesis of 2-acyl-1,3-diols bearing two stereogenic centers by the phosphine oxide-catalyzed asymmetric double aldol reaction of methyl ketones with aldehydes.^6–9) It is possible that the double aldol products 3, obtained from β-alkoxy enones 1 and aldehydes 2, are cyclized in situ, affording the corresponding highly functionalized 2,3-dihydro-4-pyranones 4 with three contiguous stereogenic centers (Fig. 1). Herein, we report the asymmetric synthesis of 2,3-dihydro-4-pyranones by asymmetric double aldol reaction/cyclization cascade.

Fig. 1. Stereoselective Synthesis of 2,3-Dihydro-4-pyranones Using Phosphine Oxide-Catalyzed Double Aldol Reaction/Cyclization Cascade

First, the reaction of 4-methoxy-3-buten-2-one (1A) with benzaldehyde (2a) was performed in the presence of 10 mol% (S)-2,2′-bis(diphenylphosphino)-1,1′-binaphthyl dioxide (BINAPO) as the Lewis base catalyst¹⁰⁾ under several reaction conditions (Table 1). The presence of silicon tetrachloride (3 eq) and N,N-diisopropylethylamine (5 eq) in CH₂Cl₂ at −40°C afforded three 2,3-dihydro-4-pyranones 5a, 6a, and 7a as a diastereomeric mixture in total 68% yield, whereas the non-cyclized product 8A was not isolated (entry 1). Furthermore, the major isomer 5a showed a promising enantioselectivity.¹¹⁾ When propionitrile was used as the solvent, the enantioselectivity improved; however, the product yield decreased (entry 2). The use of a mixture of solvents (1 : 1, v/v) afforded well-balanced yields and selectivities (entry 3). The reaction of β-benzoyloxy enone 1B afforded only non-cyclized adduct 8B in a moderate yield and with good enantioselectivity (entry 4). The reaction of β-phenoxy enone 1C resulted in a similar yield and selectivity as 1A, accompanied with non-cyclized adduct 8C (entry 5). A screening of quenching reagents showed that the prior addition of methanol (3 mL) before 1.5 M KF/3.0 M HCOOH solution promoted the cyclization of 8C, affording the corresponding 2,3-dihydro-4-pyranones in 72% yield (entry 6).¹²⁾ When the reaction was conducted at a lower concentration of 0.03 M in CH₂Cl₂–EtCN (1 : 4, v/v), the enantioselectivity increased to 91% enantiomeric excess (ee) (entry 7).

Table 1. Optimization of Reaction Condition^a)

Entry	1	Solvent	Time (h)	Yieldb) (%)	dr, 5a/6a/7ac)	ee of 5ad)
1	1A	CH2Cl2	4	68	92/2/6	57
2	1A	EtCN	4	39	92/1/7	87
3	1A	CH2Cl2–EtCN=1 : 1	4	60	85/2/13	74
4	1B	CH2Cl2–EtCN=1 : 1	2	0 (52)e)	—	— (81)e)
5	1C	CH2Cl2–EtCN=1 : 1	6	58 (22)e)	79/3/18	74
6f,g)	1C	CH2Cl2–EtCN=1 : 1	24	72	82/3/15	83
7f,g,h)	1C	CH2Cl2–EtCN=1 : 4	48	69	85/1/14	91

a) Unless otherwise noted, the reactions were carried out by adding SiCl₄ (3.0 eq) to a solution of ketone 1 (0.5 mmol), aldehyde 2a (2.5 eq), ⁱPr₂NEt (5.0 eq), and (S)-BINAPO (10 mol%) in a solvent (5 mL) at −40°C, and quenched with aq. 1.5 M KF/3.0 M HCOOH (5 mL). b) Combined yields of diastereomers. c) Determined by ¹H-NMR and HPLC analyses. d) Determined by HPLC analysis. e) Results of noncyclized adduct 8 in the parenthesis. f) The reaction was conducted at −60°C. g) The reaction was quenched with methanol (3 mL) before the addition of aq. 1.5 M KF/3.0 M HCOOH. h) The reaction was conducted at a concentration of 0.03 M.

The probable stereochemical course of the reaction is shown in Fig. 2. The double aldol reaction of enone 1 with aldehyde 2 affords two types of double aldolates, 9 and 10. Isomer 9 is optically active, whereas isomer 10 is optically inactive because of its meso-structure. It is predicted that chelation occurs between one silyloxy group and the carbonyl group, leading to the cyclization of the other silyloxy group, which is located closer to the β-position of the enone, through path A and affording the major isomer 5. When the reaction proceeds via path B, the minor isomer 6 is obtained in almost the same enantioselectivity as the major isomer 5.¹³⁾ meso-Isomer 10 reacts through path C and D equally, affording isomer 7 as a racemate.¹⁴⁾

Fig. 2. Proposed Stereochemical Course

With the optimized reaction conditions, the reactions of enone 1C with various aldehydes 2 were conducted (Table 2). Regardless of the electronic nature of the phenyl ring, similar yields and enantioselectivities of 5 were obtained (entries 2, 3). The use of p-anisaldehyde (2b) bearing an electron-donating group decreased the diastereoselectivity because of the generation of the meso-isomer in the double aldol reaction. The reactions of naphthaldehydes 2d and 2e afforded the corresponding products with good enantioselectivities (entries 4, 5).

Table 2. Reaction of Various Aldehydes^a)

Entry	2	Yieldb) (%)	dr, 5/6/7c)	ee of 5d)
1	2a, R=Ph	69	85/1/14	91
2	2b, R=4-Me-OC6H4	71	72/5/23	93
3	2c, R=4-Br-C6H4	62	81/4/15	90
4	2d, R=1-Naphthyl	64	80/5/15	77
5	2e, R=2-Naphthyl	75	84/2/14	84

a) All the reactions were carried out by adding silicon tetrachloride (3.0 eq) to a solution of ketone 1C (0.5 mmol), an aldehyde 2 (2.5 eq), N,N-diisopropylethylamine (5.0 eq), and (S)-BINAPO (10 mol%) in CH₂Cl₂–EtCN (16.7 mL, 1 : 4, v/v) at −60°C. b) Combined yields of diastereomers. c) Determined by ¹H-NMR and HPLC analyses. d) Determined by HPLC analysis.

In conclusion, we have demonstrated asymmetric 2,3-dihydro-4-pyranone synthesis using a chiral phosphine oxide as the catalyst. The chiral phosphine oxide effectively facilitated the double aldol reaction of 4-methoxy-3-buten-2-one with aldehydes and subsequent stereoselective cyclization afforded the corresponding highly functionalized 2,3-dihydro-4-pyranones in good yields and with good diastereo- and enantioselectivities. This method offers an alternative to the existing methods for the synthesis of 2,3-dihydro-4-pyranones from the corresponding ketones and aldehydes.

Experimental

General

All the reactions were performed under argon atmosphere using dried glasswares equipped with a rubber septum and a magnetic stirring bar. The column chromatography purifications were performed using Kanto Chemical Silica Gel 60N (spherical, neutral, 63–210 µm). The IR spectra were recorded using a JEOL JIR 6500-W. The ¹H- and ¹³C-NMR spectra were recorded in CDCl₃ using a JEOL JNM-ECX-400 spectrometer. The chemical shifts are reported in ppm relative to the internal tetramethylsilane (TMS) standard (δ 0.00 ppm) for the ¹H-NMR spectra and solvent signals (δ 77.0 ppm) for the ¹³C-NMR spectra. The mass spectra were recorded using a JEOL JMS-700MStation mass spectrometer. The HPLC analyses were performed using JASCO P-2080 and UV-2075.

Typical Procedure for the Synthesis of 2,3-Dihydro-4-pyranones

Silicon tetrachloride (3 eq) was added to a solution of (S)-BINAPO (10 mol%), ketone 1C (0.5 mmol), aldehyde 2a (2.5 eq), and N,N-diisopropylethylamine (5 eq) in a mixture of CH₂Cl₂–EtCN (1 : 4, v/v, 16.7 mL) at −60°C and the reaction mixture was stirred for 24 h at the same temperature. MeOH (3 mL) was added to the resulting solution. After 1 h, aq. 1.5 M KF/3.0 M HCOOH (5 mL) was added to the resulting solution. The slurry was stirred for 1 h at room temperature, and the two-layer mixture was extracted with EtOAc (3×15 mL). The combined organic layer was washed with 10% HCl, sat. NaHCO₃, and brine and dried over Na₂SO₄. After the filtration and concentration, the crude product was purified by column chromatography, affording products 5a, 6a, and 7a as a diastereomeric mixture.

(2S,3R,1′R)-3-[Hydroxy(phenyl)methyl]-2-phenyl-2,3-dihydro-4-pyranone

Data for 5a

TLC: Rf 0.57 (hexane–EtOAc=1 : 1, stained orange with anisaldehyde); [α]_D³¹ −195.3 (c=1.0, CHCl₃) for 99% ee (after recrystallization); ¹H-NMR (CDCl₃) δ: 3.50 (1H, dd, J=5.0, 10.6 Hz), 4.30 (1H, d, J=7.8 Hz), 4.89 (1H, dd, J=5.0, 7.8 Hz), 5.30 (1H, d, J=10.6 Hz), 5.53 (1H, d, J=6.0 Hz), 7.10–7.34 (11H, m); ¹³C-NMR (CDCl₃) δ: 54.4, 72.9, 82.0, 107.6, 126.9, 127.8, 127.8, 128.3, 128.7, 129.3, 135.7, 140.6, 163.1, 194.4; IR (film): 1599, 1672, 3444 cm⁻¹; low resolution (LR)-MS (FAB): m/z 303 (M+Na)⁺; high resolution (HR)-MS (FAB): Calcd for C₁₈H₁₆O₃Na 303.0997. Found 303.1002. The ee was determined to be 91% ee using HPLC equipped with a Daicel Chiralpak AD-H column [eluent: hexane–IPA=9 : 1; flow rate: 1.0 mL/min; detection: 254 nm; t_R: 15.6 min (major), 17.7 min (minor)].

Data for 6a

TLC: Rf 0.55 (hexane–EtOAc=1 : 1, stained orange with anisaldehyde); ¹H-NMR (CDCl₃) δ: 3.49 (1H, dd, J=6.0, 8.2 Hz), 4.38 (1H, br s), 4.91 (1H, d, J=8.2 Hz), 5.04 (1H, d, J=6.0 Hz), 5.58 (1H, d, J=6.0 Hz), 6.97–6.99 (2H, m), 7.21–7.25 (6H, m), 7.31–7.39 (3H, m); ¹³C-NMR (CDCl₃) δ: 54.1, 72.3, 81.4, 107.0, 126.8, 128.4, 128.4, 128.6, 128.7, 129.4, 134.6, 139.4, 162.0, 195.7; IR (film): 1599, 1653, 3444 cm⁻¹; LR-MS (FAB): m/z 303 (M+Na)⁺; HR-MS (FAB): Calcd for C₁₈H₁₆O₃Na 303.0997. Found 303.1009.

Data for 7a

TLC: Rf 0.49 (hexane–EtOAc=1 : 1, stained orange with anisaldehyde); ¹H-NMR (CDCl₃) δ: 3.15 (1H, d, J=6.4 Hz), 3.27 (1H, dd, J=6.0, 7.3 Hz), 4.83 (1H, dd, J=6.0, 6.4 Hz), 5.39 (1H, d, J=7.3 Hz), 5.49 (1H, d, J=6.0 Hz), 7.26–7.38 (11H, m); ¹³C-NMR (CDCl₃) δ: 56.3, 72.7, 82.0, 107.0, 126.2, 127.0, 128.0, 128.6, 128.8, 129.0, 136.6, 141.4, 162.1, 193.5; IR (film): 1599, 1651, 3408 cm⁻¹; LR-MS (FAB): m/z 303 (M+Na)⁺; HR-MS (FAB): Calcd for C₁₈H₁₆O₃Na 303.0997. Found 303.1000.

3-[Hydroxyl(4-methoxyphenyl)methyl]-2-(4-methoxyphenyl)-2,3-dihydro-4-pyranone

Data for 5b

TLC: Rf 0.22 (hexane–EtOAc=3 : 2, stained orange with anisaldehyde); [α]_D³³ −330.1 (c=1.1, CHCl₃) for 99% ee (after recrystallization); ¹H-NMR (CDCl₃) δ: 3.54 (1H, dd, J=4.8, 11.9 Hz), 3.79 (3H, s), 3.83 (3H, s), 4.64 (1H, d, J=8.7 Hz), 4.66 (1H, dd, J=4.8, 8.7 Hz), 5.08 (1H, d, J=11.9 Hz), 5.52 (1H, d, J=6.0 Hz), 6.79 (2H, d, J=8.7 Hz), 6.89 (2H, d, J=8.7 Hz), 7.03 (2H, d, J=8.7 Hz), 7.21 (2H, d, J=8.7 Hz), 7.35 (1H, d, J=6.0 Hz); ¹³C-NMR (CDCl₃) δ: 53.9, 53.3, 55.4, 73.2, 82.3, 107.8, 113.7, 114.2, 127.6, 128.6, 129.6, 132.7, 159.3, 160.5, 163.5, 195.8; IR (film): 1252, 1597, 1659, 3425 cm⁻¹; LR-MS (FAB): m/z 363 (M+Na)⁺; HR-MS (FAB): Calcd for C₂₀H₂₀O₅Na 363.1208. Found 363.1213. The ee was determined to be 93% ee using HPLC equipped with Daicel Chiralcel OZ-H column [eluent: hexane–IPA=3 : 1; flow rate: 1.0 mL/min; detection: 254 nm; t_R: 18.5 min (major), 23.1 min (minor)].

3-[(4-Bromophenyl)hydroxymethyl]-2-(4-bromophenyl)-2,3-dihydro-4-pyranone

Data for 5c

TLC: Rf 0.29 (hexane–EtOAc=2 : 1, stained orange with anisaldehyde); [α]_D³¹ −156.4 (c=1.0, CHCl₃) for 99% ee (after recrystallization); ¹H-NMR (CDCl₃) δ: 3.71 (1H, d, J=4.1, 11.2 Hz), 3.74 (1H, d, J=7.4 Hz), 5.10 (1H, dd, J=4.1, 7.4 Hz), 5.30 (1H, d, J=11.2 Hz), 5.57 (1H, d, J=6.0 Hz), 6.91 (1H, d, J=8.7 Hz), 7.07 (1H, d, J=8.7 Hz), 7.32 (1H, d, J=8.3 Hz), 7.39 (1H, d, J=6.0 Hz), 7.42 (1H, d, J=8.3 Hz); ¹³C-NMR (CDCl₃) δ: 54.7, 71.1, 81.4, 107.6, 121.5, 123.6, 128.1, 129.6, 131.3, 131.8, 134.5, 139.7, 163.2, 193.4; IR (film): 1595, 1662, 3410 cm⁻¹; LR-MS (FAB): m/z 437, 439, 441 (M+H)⁺; HR-MS (FAB): Calcd for C₁₈H₁₅Br₂O₃ 436.9368. Found 436.9351. The ee was determined to be 90% ee using HPLC equipped with a Daicel Chiralpak AD-H column [eluent: hexane–IPA=9 : 1; flow rate: 0.5 mL/min; detection: 254 nm; t_R: 28.2 min (major), 41.0 min (minor)].

3-[Hydroxyl(1-naphthyl)methyl]-2-(1-naphthyl)-2,3-dihydro-4-pyranone

Data for 5d

TLC: Rf 0.35 (hexane–EtOAc=2 : 1, stained orange with anisaldehyde); [α]_D³⁶ −33.1 (c=1.6, CHCl₃) for 77% ee; ¹H-NMR (CDCl₃) δ: 2.55 (1H, d, J=4.6 Hz), 3.74 (1H, dd, J=2.8, 7.8 Hz), 5.70 (1H, d, J=6.0 Hz), 6.34 (1H, m), 6.55 (1H, d, J=7.8 Hz), 7.03–7.92 (15H, m); ¹³C-NMR (CDCl₃) δ: 54.1, 69.0, 78.0, 107.1, 122.4, 122.8, 123.6, 124.4, 124.6, 125.3, 125.4, 125.9, 126.2, 127.7, 128.6, 128.7, 129.3, 130.2, 131.9, 133.3, 133.6, 136.2, 163.1, 192.8; IR (film): 796, 1597, 1657, 3417 cm⁻¹; LR-MS (FAB): m/z 380 (M)⁺; HR-MS (FAB): Calcd for C₂₆H₂₀O₃ 380.1412. Found 380.1411. The ee was determined to be 77% ee using HPLC equipped with a Daicel Chiralpak AS-H column [eluent: hexane–IPA=9 : 1; flow rate: 1.0 mL/min; detection: 254 nm; t_R: 21.4 min (major), 35.5 min (minor)].

3-[Hydroxyl(2-naphthyl)methyl]-2-(2-naphthyl)-2,3-dihydro-4-pyranone

Data for 5e

TLC: Rf 0.35 (hexane–EtOAc=2 : 1, stained orange with anisaldehyde); [α]_D³⁰ −182.2 (c=1.0, CHCl₃) for 84% ee; ¹H-NMR (CDCl₃) δ: 3.71 (1H, dd, J=4.6, 11.4 Hz), 4.30 (1H, d, J=7.8 Hz), 5.17 (1H, dd, J=4.6, 7.8 Hz), 5.45 (1H, d, J=11.4 Hz), 5.61 (1H, d, J=6.0 Hz), 7.28–7.76 (15H, m); ¹³C-NMR (CDCl₃) δ: 54.1, 72.8, 82.4, 107.7, 124.3, 124.4, 125.9, 126.1, 126.2, 126.5, 126.8, 127.4, 127.6, 127.8, 128.0, 128.1, 128.2, 128.6, 132.7, 132.8, 133.5, 137.9, 163.3, 194.6; IR (film): 748, 816, 856, 1597, 1652, 3404 cm⁻¹; LR-MS (FAB): m/z 403 (M+Na)⁺; HR-MS (FAB): Calcd for C₂₆H₂₀O₃Na 403.1310. Found 403.1312. The ee was determined to be 84% ee using HPLC equipped with a Daicel Chiralpak AD-H column [eluent: hexane–IPA=9 : 1; flow rate: 1.0 mL/min; detection: 254 nm; t_R: 27.6 min (major), 30.6 min (minor)].

Acknowledgments

This work was partially supported by JSPS KAKENHI (Grant No. 25860008) and a Grant-in-Aid for Scientific Research on Innovative Areas “Advanced Molecular Transformations by Organocatalysts” from The Ministry of Education, Culture, Sports, Science and Technology of Japan.

Conflict of Interest

The authors declare no conflict of interest.

References and Notes

• 1) Clarke P. A., Santos S., Eur. J. Org. Chem., 2006, 2045–2053 (2006).
• 2) Pellissier H., Tetrahedron, 65, 2839–2877 (2009).
• 3) Crane E. A., Scheidt K. A., Angew. Chem. Int. Ed., 49, 8316–8326 (2010).
• 4) Olier C., Kaafarani M., Gastaldi S., Bertrand M. P., Tetrahedron, 66, 413–445 (2010).
• 5) Han X., Peh G. R., Floreancig P. E., Eur. J. Org. Chem., 2013, 1193–1208 (2013).
• 6) Leading paper for boron triflate-mediated double aldol reaction, see: Abiko A., Liu J.-F., Buske D. C., Moriyama S., Masamune S., J. Am. Chem. Soc., 121, 7168–7169 (1999).
• 7) Shimoda Y., Kotani S., Sugiura M., Nakajima M., Chem. Eur. J., 17, 7992–7995 (2011).
• 8) Kotani S., Sugiura M., Nakajima M., Synlett, 25, 631–640 (2014).
• 9) Shimoda Y., Kubo T., Sugiura M., Kotani S., Nakajima M., Angew. Chem. Int. Ed., 52, 3461–3464 (2013).
• 10) Denmark S. E., Beutner G. L., Angew. Chem. Int. Ed., 47, 1560–1638 (2008).
• 11) The absolute configuration of the major isomer 5a was determined to be 2S,3R,1′R by X-ray crystallographic analysis after derivatization to the corresponding camphorsulfonate. The crystallographic data for this structure have been deposited with the Cambridge Crystallographic Data Centre as a supplementary publication number CCDC 1425234.
• 12) The treatment of trichlorosilyl aldolate 9 with MeOH appeared to afford trimethoxysilyl aldolate, which could be transformed to 5a.
• 13) Minor isomer 6a has almost the same enantioselectivity as the major isomer 5a.
• 14) It is possible to generate two types of meso-isomers in the double aldol reaction. But, one cyclized isomer was isolated after cyclization. The relative configuration of 7a was assigned on the basis of the proton–proton NMR coupling constants.