Introduction

Piperidines are among the nitrogen heterocycles most frequently used in pharmaceuticals [1]. Apart from their appearance in pharmaceutically interesting substances [2], tetrahydropyridines, and piperidines are present as common structural motifs in many bioactive natural products [3,4,5]. Because of the interesting biological effects exerted by tetrahydropyridine- and piperidine-containing compounds, many methods have been developed to access these valuable structures [2, 6]. In 2003, we disclosed that functionalized tetrahydropyridines could be synthesized through a PBu3-catalyzed [4 + 2] reaction of allenoates and imines [7, 8]. Subsequently, Fu [9], Zhao [10], Guo [11], and Sasai [12] realized asymmetric versions of the [4 + 2] reaction using various chiral phosphines. In the last 50 years, major efforts have been exerted in the development of chiral phosphine ligands [13,14,15]. The chirality of a phosphine may reside on the phosphorus center, in the carbon backbone, or in the phosphine’s molecular framework [16]. As part of our efforts in phosphine organocatalysis, we have developed a class of structurally unique P-chiral [2.2.1] bicyclic phosphines [17, 18]. Recently, we applied one such bridged bicyclic P-chiral phosphine, exo-(p-anisyl)-HypPhos, in the asymmetric allene–imine [4 + 2] annulation to afford valuable chiral piperidines in good yields and with excellent enantioselectivities [19]. Herein, we report a class of new chiral phosphines derived from l-hydroxyproline (l-Hyp) and their application in the allene–imine [4 + 2] annulation.

The abundance and ready availability of chiral α-amino acids have made them desirable sources of chirality. Indeed, many bifunctional chiral aminophosphines have been derived from α-amino acids [14]. In previous studies, the presence of a free NH group, responsible for hydrogen bonding, has been the major focus when designing chiral aminophosphines for asymmetric phosphine organocatalysis (Scheme 1a) [14]. Progress in the field of enamine and iminium catalysis has revealed that steric interaction is another useful mode of asymmetric induction [20,21,22,23]. Nevertheless, steric interaction has not previously been considered a promising platform for the design of α-amino acid-derived aminophosphines. l-Hyp, an abundant nonproteinogenic amino acid having a proline skeleton and an additional OH group for further steric tuning, has been chosen as the platform for our catalyst design (Scheme 1a). In addition to introducing sterically tunable protecting groups at the pyrroline nitrogen atom and the 4-OH group, a hydrogen-bonding site could also be installed—for example, as a urea or thiourea motif—at the pyrroline nitrogen atom. We envisioned that the aminophosphines 2af could be obtained from l-Hyp by installing a diphenylphosphino group through the displacement of the mesylate in 1 and subsequent functional group manipulations (Scheme 1b). The synthesis of the mesylate 1 also appeared to be straightforward. Starting from commercially available l-Hyp, various alcohol and amine protecting groups could be installed prior to substitution of the mesylate by a diphenylphosphino group.

Scheme 1
scheme 1

Synthetic design of chiral aminophosphines 2 from l-Hyp

Results and discussion

The synthesis of compound 2a is illustrated in Scheme 2. First, the Boc-protection of l-Hyp proceeded to provide N-Boc-l-Hyp 3. Benzylation of the free OH group followed by borane-mediated reduction of the carboxyl group resulted in the primary alcohol 5. Only one column chromatographic purification through SiO2 was necessary to prepare 16.3 g of 5 in 70% yield in the three-step process. Mesylation of 5 led to the key intermediate 1 for phosphine synthesis. The mesylate 1 was converted into the phosphine 2a in high yield after slow addition of an excess of potassium diphenylphosphide at low temperature.

Scheme 2
scheme 2

Preparation of the phosphine 2a

Having prepared the aminophosphine 2a, the acid-liable N-Boc protecting group was readily replaced with several acyl groups to afford the chiral aminophosphines 2bf of varying steric bias (Scheme 3). After Boc-deprotection, an excess of triethylamine was added to neutralize TFA and promote the acylation with the acyl chloride. Using this one-pot procedure, the four aminophosphines 2be were prepared in high yields. When phenyl isocyanate was used instead of an acyl chloride, the phosphine 2f, presenting a hydrogen-bonding site, was obtained in 77% yield.

Scheme 3
scheme 3

Preparation of the phosphines 2bf

The six chiral phosphines with varied N-substituents were then applied to our [4 + 2] annulation [7]. We first examined the [4 + 2] annulation of the activated allenoate 8 and the imine 7 mediated by the phosphine 2a in CH2Cl2 (Table 1, entry 1). An inseparable mixture of the tetrahydropyridine 9 and the tetrahydropyridine 10 was obtained in 68% yield, with the ee of 9 being 30%. When the solvent was switched to benzene, the ee of 9 increased to 42%, but 10 was now the major product (entry 2). We then tested the other phosphines for the annulation in CH2Cl2 (entries 3–6). When using the catalyst 2b for the annulation in CH2Cl2, the ee of 9 increased to 42% compared with that of 2a in the same solvent (cf. entries 3 and 1). Under otherwise identical conditions, the ratio of 10 had also increased. The phosphine 2d, featuring a bulkier acyl substituent on the nitrogen atom, produced the tetrahydropyridine 9 with a higher ee of 53% (entry 5). Interestingly, in the reaction catalyzed by the phosphine 2e presenting an adamantylcarbonyl group, the selectivity of the product was completely reversed—now favoring the tetrahydropyridine 10, which was isolated in 80% yield and contaminated with only a trace of 9 (entry 6). The NH group in the phosphine 2f seemed to have little effect on the enantioselectivity, but the formation of 10 was largely inhibited (entry 7). From these preliminary results, it appeared that the enantioselectivity increased upon increasing the bulk of the N-acyl substituent, presumably through a mechanism involving steric control.

Table 1 Phosphines 2–catalyzed [4 + 2] annulation of the allenoate 8a

It was also interesting that more sterically demanding phosphines could facilitate the formation of 10 over 9. The reversed regioselectivity can be explained by two possible pathways of the imine–allene [4 + 2] reaction (Scheme 4) [7, 19, 24]. At the onset of the [4 + 2] reaction, addition of the phosphine 2 to the allene 8 generates intermediate A, which is in resonance with structure B, which, in turn, is in equilibrium with intermediate B′. In path I, γ-addition of intermediate B to the imine 7 affords intermediate C, followed by the formation of 9. In path II, β′-addition of intermediate B′ to the imine 7 generates intermediate C′, followed by the formation of 10. When a bulkier phosphine was used in the reaction, the γ-position of intermediate B was more encumbered, relative to its β′-position. In this case, path II, which afforded the product 10, became the favored pathway.

Scheme 4
scheme 4

Possible reaction pathways for the allene–imine [4 + 2] annulation

The same catalysts were then employed in the allene–imine [4 + 2] annulation of the less activated α-methylallenoate 11 (Table 2). Because the reaction was slower with this allene 11, the catalyst loading was increased to 30 mol% and 4-Å MS were added to minimize hydrolysis of the imine 7. Although the conversion of the imine 7 and the yield of 12 were not good, we observed a similar trend in the induction of enantioselectivity as that in Table 1. Increasing the bulk of the N-acyl substituent in the phosphine 2 afforded greater enantioselectivity for the formation of the tetrahydropyridine 12. When using the catalyst 2d, an ee of 41% was induced (entry 4). The bulky phosphines again facilitated the formation of the β′-addition product, in this case 13 (cf. entries 4 vs. 1 and 2). These preliminary findings, obtained from nonoptimized reaction conditions, were consistent with the model of steric control.

Table 2 Phosphines 2–catalyzed [4 + 2] annulation of the α-methylallenoate 11a

Conclusion

Taking advantage of the ready availability of naturally occurring chiral α-amino acids, we have chosen l-Hyp, a nonproteinogenic—but abundantly available—amino acid, for the design of several chiral phosphines. With the goal of exploiting steric interactions for enantioinduction, we have developed a synthetic route toward a number of l-Hyp–derived chiral phosphines. The preliminary application of these phosphines in the allene−imine [4 + 2] annulation has proven moderately successful in terms of enantioselectivities. When using bulky phosphines, an interesting switch of the reaction site occurred: from the γ- to the β′-carbon atom of the allenoate. The clear trend in the enantioinduction hints at the potential of developing chiral phosphines that operate with steric tuning as the sole source of enantioselectivity control in the [4 + 2] annulation.

Experimental

4-Benzyloxy-2-methanesulfonyloxymethyl-pyrrolidine-1-carboxylic acid tert-butyl ester (1)

Boc2O (18.3 g, 83.9 mmol) was added slowly to a solution of l-Hyp (10.0 g, 76.3 mmol) in 1,4-dioxane (80 mL) and 1 N aqueous NaOH (80 mL) at 0 °C. The reaction mixture was stirred for 8 h at room temperature, then concentrated under reduced pressure, acidified to pH 1 using 2 N aqueous HCl, and extracted with EtOAc. The extract was washed with brine, dried (Na2SO4), and concentrated to provide the Boc-protected l-Hyp 3 [25] as a clear viscous oil. This crude product was used in the next step without further purification.

The Boc-protected l-Hyp 3 was dissolved in THF (100 mL) and then cannulated into a slurry of NaH (60% in mineral oil, 11.2 g, 155 mmol) in THF (200 mL) at 0 °C. Benzyl bromide (15.0 g, 87.7 mmol) was added dropwise to the reaction mixture. The flask was warmed to room temperature and then the mixture was heated under reflux for 6 h. The reaction mixture was cooled to room temperature and poured over ice. The organic solvent was evaporated under reduced pressure and the aqueous solution was washed with EtOAc. The aqueous phase was acidified with 2N HCl until the pH was 2, and then it was extracted with EtOAc. The organic phase was concentrated under reduced pressure to yield 4 as a brownish yellow oil. This brownish yellow oil was dissolved in dry THF (300 mL) and cooled to 0 °C. BH3·DMS (5.7 mL, 58 mmol) was added dropwise to the reaction mixture, which was then kept stirring at 0 °C for an additional 1 h. The flask was removed from the ice bath, and the mixture stirred overnight at room temperature. The reaction mixture was poured over ice and sequentially extracted with EtOAc, washed with saturated aqueous NaHCO3, washed with brine, and dried (Na2SO4). The solution was concentrated under reduced pressure and purified through flash column chromatography (FCC; gradient EtOAc/hexanes, from 30 to 50%) to yield 5 [26] as a slightly yellow oil (16.3 g, 70% over three steps).

The slightly yellow oil from the previous step (16.1 g, 52.3 mmol) was dissolved in dry CH2Cl2 (160 mL) and cooled to 0 °C. Et3N (7.8 mL, 56.0 mmol) was added and then MsCl (6.3 g, 54.9 mmol) was added slowly. After the reaction had reached completion (checked using TLC), the solvent was evaporated under reduced pressure. The residue was dissolved in EtOAc and washed sequentially with water, saturated aqueous NaHCO3, and brine, then dried (Na2SO4). The product was isolated through FCC (EtOAc/hexanes, 30%) to provide 1 [26] (20.0 g, 99%) as a slightly yellow viscous oil.

(2S,4R)-tert-butyl 4-(benzyloxy)-2-((diphenylphosphino)methyl)pyrrolidine-1-carboxylate (2a)

Compound 1 (771 mg, 2 mmol) was placed in a flame-dried round-bottom flask containing a stirrer bar. Dry THF (20 mL) was added via syringe under argon. The flask was cooled to −45 °C (MeCN/dry ice bath) and then 0.5 N Ph2PK solution in THF (6 mL, 3 mmol) was added over 2 h via syringe pump. When the addition was complete, the mixture was stirred at the same temperature for an additional 7 h. The flask was warmed to room temperature and then H2O (1–1.5 mL) was added until the reaction color disappeared. CH2Cl2 (20 mL) was added and then the organic phase was collected. The aqueous phase was extracted twice with CH2Cl2. The combined organic phases were dried (Na2SO4) and concentrated under reduced pressure. The residue was purified through FCC (EtOAc/hexanes, from 1:10 to 1:5) to afford 2a (846 mg, 89%) as a clear viscous oil. \([{\rm{\alpha}}]^{26}_{\mathrm{D}}\) –47.4 (с 0.35, acetone); 1H NMR (500 MHz, CDCl3) δ 7.61 (br s, 1H), 7.54 (br s, 1H), 7.46 (app t, J = 6.6 Hz, 2H), 7.38–7.30 (m, 11H), 4.49 (br s, 2H), 4.27–4.10 (m, 2H), 3.90–3.43 (m, 2H), 3.04–2.88 (m, 1H), 2.33–2.12 (m, 3H), 1.49 (s, 9H); 13C NMR (125 MHz, CDCl3) δ 154.4, 137.8, 132.7 (d, J = 19.2Hz), 132.3 (d, J = 19.5 Hz), 128.6, 128.3 (d, J = 7.1 Hz), 128.2, 127.4, 127.3, 79.4, 70.6, 54.1 (d, J = 18.9 Hz), 53.2, 38.6, 35.0 (d, J = 10.8 Hz), 33.5 (br s), 28.3; 31P NMR (202 MHz, CDCl3) δ −22.9; IR (film) νmax 2975, 2930, 2362, 2338, 1690, 1393, 1171, 739, 697 cm−1; HRMS-DART: calcd for C29H35NO3P ([M+H]+) m/z 476.2349; found 476.2345.

General procedure for preparing 2be

The phosphine 2a (0.20 g, 0.42 mmol) was dissolved in CH2Cl2 (1.5 mL) in a flame-dried round-bottom flask containing a stirrer bar. TFA (0.5 mL) was added dropwise to the solution at 0 °C and then the mixture was stirred at room temperature for 1 h. CH2Cl2 (3 mL) was added and then the mixture was cooled to 0 °C. NEt3 (1.2 mL) was added dropwise, and then an acyl chloride (1.2 eq) was added slowly. The mixture was stirred at room temperature overnight and then washed with H2O. The organic phase was dried (Na2SO4) and concentrated under reduced pressure. The residue was purified through FCC.

((2S,4R)-4-(Benzyloxy)-2-((diphenylphosphino)methyl)pyrrolidin-1-yl)(3,5-bis(trifluoromethyl)phenyl)methanone (2b)

Obtained as a white solid (70%). \([{\rm{\alpha}}]^{24}_{\mathrm{D}}\) –58.77 (с 0.65, acetone); 1H NMR (500 MHz, CDCl3) δ 7.87 (s, 1H), 7.60 (s, 2H), 7.57–7.53 (m, 4H), 7.38–7.16 (m, 11H), 4.78–4.70 (m, 1H), 4.46 (d, J = 11.9Hz, 1 H), 4.30 (d, J = 11.9 Hz, 1H), 4.14 (br s, 1H), 3.53 (dd, J = 11.3, 3.3 Hz, 1H), 3.32 (d, J = 11.3 Hz, 1H), 2.84–2.76 (m, 2H), 2.48 (dd, J = 13.2, 7.9 Hz, 1H), 2.34 (br s, 1H); 13C NMR (125 MHz, CDCl3) δ 167.1, 138.4, 137.6, 137.5, 137.49, 132.8 (d, J = 21.3 Hz), 132.4 (d, J = 18.6 Hz), 131.5 (q, J = 33.7 Hz), 129.1, 128.63, 128.56 (d, J = 7.5 Hz), 128.54 (d, J = 6.8 Hz), 128.4, 127.7, 127.6 (br s), 127.4, 123.6 (sp, J = 3.6 Hz), 122.8 (q, J = 273.7 Hz), 77.1, 70.8, 55.6, 55.0 (d, J = 16.9 Hz), 36.9 (d, J = 10.6 Hz), 32.3 (d, J = 12.9 Hz); 31P NMR (202 MHz, CDCl3) δ –26.4; IR (film) νmax 3057, 2926, 2867, 1642, 1363, 1280, 1137, 906, 740, 697 cm−1; HRMS-DART: calcd for C33H29F6NO2P ([M+H]+) m/z 616.1830; found 616.1830.

1-((2S,4R)-4-(benzyloxy)-2-((diphenylphosphino)methyl)pyrrolidin-1-yl)-2,2-dimethylpropan-1-one (2c)

Obtained as a white solid (78%). \([{\rm{\alpha}}]^{26}_{\mathrm{D}}\) –4.0 (с 0.75, acetone); 1H NMR (500 MHz, CDCl3) δ 7.60 (t, J = 7.5 Hz, 2H), 7.44–7.41 (m, 2H), 7.37 (app t, J = 7.0 Hz, 2H), 7.34–7.26 (m, 9H), 4.52–4.43 (m, 3H) (apparent overlapping peaks), 4.17 (app s, 1H), 3.90 (d, J = 11.4 Hz, 1H), 3.56 (dd, J = 11.4, 4.4 Hz, 1 H), 2.91 (d, J = 13.9 Hz, 1H), 2.24–2.19 (m, 2 H), 2.16–2.10 (m, 1H), 1.17 (s, 9H); 13C NMR (125 MHz, CDCl3) δ 176.7, 137.8, 132.8 (d, J = 18.8 Hz), 132.7 (d, J = 19.4 Hz), 128.6, 128.5 (d, J = 7.0 Hz), 128.35 (d, J = 7.0 Hz), 128.32, 127.6, 127.3, 77.4, 70.6, 55.7 (d, J = 17.5 Hz), 53.1, 38.8, 35.5 (d, J = 9.3 Hz), 32.8 (d, J = 13.7 Hz), 27.5; 31P NMR (202 MHz, CDCl3) δ –23.5; IR (film) νmax 2969, 2930, 1620, 1407, 1093, 1074, 738, 697 cm–1; HRMS-DART: calcd for C29H35NO2P ([M+H]+) m/z 460.2400; found 460.2397.

1-((2S,4R)-4-(benzyloxy)-2-((diphenylphosphino)methyl)pyrrolidin-1-yl)-2,2-diphenylethanone (2d)

Obtained as a pale-yellow solid (73%). \([{\rm{\alpha}}]^{27}_{\mathrm{D}}\) –50.14 (с 0.17, acetone); 1H NMR (500 MHz, CDCl3) δ (in aliphatic region, major rotamer) 5.00 (s, 1H), 4.52–4.50 (m, 1H) (overlapping peaks), 4.25 (d, J = 12.2 Hz, 1H), 4.17 (d, J = 11.9 Hz, 1H), 4.13–4.10 (m, 1H), 3.66 (d, J = 10.2 Hz, 1H), 3.49 (dd, J = 11.0, 4.7 Hz, 1H), 3.26 (dt, J = 13.5, 4.5 Hz, 1H), 2.24–2.14 (m, 2H), 2.05–2.00 (m, 1H); 13C NMR (125 MHz, CDCl3) δ (major rotamer) 170.3, 139.4, 138.9, 137.7, 133.0 (d, J = 19.3 Hz), 132.7 (d, J = 19.3 Hz), 129.1, 128.9, 128.8, 128.6, 128.59, 128.56, 128.43, 128.41, 128.35, 128.32, 127.6, 127.5, 126.9 (d, J = 8.0 Hz), 76.5 (d, J = 1.8 Hz), 70.7, 57.0, 54.9 (d, J = 19.2 Hz), 51.9, 36.9 (d, J = 9.8 Hz), 32.7 (d, J = 12.5 Hz); 31P NMR (202 MHz, CDCl3) δ –22.5 (major rotamer), –23.2 (minor rotamer); IR (film) νmax 3060, 3028, 1645, 1411, 1192, 738, 697 cm–1; HRMS-DART: calcd for C38H37NO2P ([M+H]+) m/z 570.2556; found 570.2554.

(1S,3R)-adamantan-1-yl((2S,4R)-4-(benzyloxy)-2-((diphenylphosphino)methyl)pyrrolidin-1-yl)methanone (2e)

Obtained as a colorless sticky compound (98%). \([{\rm{\alpha}}]^{27}_{\mathrm{D}}\) + 2.7 (с 1.20, acetone); 1H NMR (500 MHz, CDCl3) δ 7.60 (t, J = 7.3 Hz, 2H), 7.44–7.41 (m, 2H), 7.39–7.28 (m, 11H), 4.54–4.50 (m, 2H) (overlapping peaks), 4.44 (d, J = 12 Hz, 1H), 4.19 (br s, 1H), 3.99 (d, J = 11.2 Hz, 1H), 3.62 (dd, J = 11.3, 4.4 Hz, 1H), 2.88 (d, J = 13.3 Hz, 1H), 2.19 (dd, J = 13.4, 9.5 Hz, 2H), 2.12 (br s, 1H), 1.99 (s, 3H), 1.87 (s, 6H), 1.71–1.65 (app q, J = 12.4 Hz, 6H); 13C NMR (125 MHz, CDCl3) δ 176.1, 137.9, 132.8 (d, J = 18.8 Hz), 132.7 (d, J = 19.6 Hz), 128.53, 128.48, 128.34, 128.30, 127.6, 127.5, 77.6, 70.8, 55.9 (d, J = 16.3 Hz), 53.1, 41.6, 38.0, 36.5, 35.2 (d, J = 8.2 Hz), 32.9 (d, J = 12.0 Hz), 28.3; 31P NMR (202 MHz, CDCl3) δ –21.8; IR (film) νmax 3053, 2905, 2850, 2235, 1610, 1391, 1073, 909, 734, 697 cm–1; HRMS-DART: calcd for C35H41F6NO2P ([M + H]+) m/z 538.2869; found 538.2872.

(2S,4R)-4-(benzyloxy)-2-((diphenylphosphanyl)methyl)-N-phenylpyrrolidine-1-carboxamide (2f)

The phosphine 2a (76.0 mg, 0.158 mmol) was dissolved in CH2Cl2 (0.5 mL) in a flame-dried round-bottom flask containing with a stirred bar. TFA (0.17 mL) was added dropwise to the solution at 0 °C and then the mixture was stirred at room temperature for 1 h. CH2Cl2 (1 mL) was added and then the mixture was cooled to 0 °C. NEt3 (0.4 mL) was added dropwise and then the reaction was quenched through the addition of H2O (2 mL). The aqueous phase was exacted with CH2Cl2. The combined organic phases were washed with brine, dried (Na2SO4), and concentrated. The residue was dissolved in CH2Cl2 (1 mL). Phenyl isocyanate (21.0 mg, 0.176 mmol) was added and then the mixture was stirred overnight. The solution was concentrated and the residue purified through FCC (hexanes/EtOAc, 5:1) to give 2f as a white semisolid. \([{\rm{\alpha}}]^{28}_{\mathrm{D}}\) –39.8 (с 1.00, acetone); 1H NMR (500 MHz, d6-acetone) δ 8.12 (s, 1H), 7.66–6.87 (m, 20H), 4.47 (s, 2H), 4.37–4.21 (m, 2H), 3.68 (d, J = 4.1 Hz, 2H), 3.05 (dt, J = 13.5, 3.4 Hz, 1H), 2.28–2.20 (m, 1H), 2.18–2.06 (m, 2H). 13C NMR (100 MHz, d6-acetone) δ 154.4, 140.5, 139.6 (d, J = 12.9 Hz), 138.7, 138.2 (d, J = 13.0 Hz), 132.8 (d, J = 19.6 Hz), 132.6 (d, J = 20.0 Hz), 128.7, 128.6 (d, J = 2.5 Hz), 128.5, 128.4 (d, J = 7.0 Hz), 128.3, 128.25 (d, J = 8.1 Hz), 127.4, 121.9 (d, J = 16.7 Hz), 119.5, 118.5, 77.0 (d, J = 1.7 Hz), 70.4, 54.70 (d, J = 20.4 Hz), 51.9, 37.3 (d, J = 8.4 Hz), 33.7 (d, J = 13.9 Hz). 31P NMR (121 MHz, d6-acetone) δ –23.12. IR (film) νmax 3324, 3054, 2926, 2426, 1700, 1633, 1443, 1385, 1246, 750, 695 cm−1; HRMS-DART: calcd for C31H32N2O2P ([M+H]+) m/z 495.2196; found 485.2219.

General experimental procedure for the allene–imine [4 + 2] annulation (with the allene 8)

The allenoate (0.12 mmol) was added in one portion to a solution of N-tosylimine (0.1 mmol) and the aminophosphine catalyst (0.02 mmol) in the solvent (2 mL). The flask was capped with a Teflon cap under a N2 flow. The mixture was stirred at room temperature for 2 days. The resulting solution was concentrated and the residue purified through FCC on silica gel (eluent: 20–30% EtOAc in hexane) to give the product.

Diethyl (2S,3S)-2-phenyl-1-tosyl-1,2,3,6-tetrahydropyridine-3,4-dicarboxylate (10)

Obtained as a white solid (80%). 57% ee—determined using an SFC instrument, a Daicel ChiralPak IC-3 column, and 20% MeCN; tR (minor) = 3.74 min; tR (major) = 4.81 min; \([{\rm{\alpha}}]^{25}_{\mathrm{D}}\) –38.0 (с 1.10, acetone); 1H NMR (400 MHz, CDCl3) δ 7.64 (d, J = 8.3 Hz, 2H), 7.28–7.16 (m, 7H), 6.94 (dd, J = 3.7, 2.7 Hz, 1H), 5.93 (s, 1H), 4.24–4.12 (m, 4H), 4.11–3.97 (m, 2H), 3.69 (dt, J = 19.9, 2.4 Hz, 1H), 2.38 (s, 3H), 1.26 (t, J = 7.4 Hz, 3H), 1.22 (t, J = 7.3 Hz, 3H); 13C NMR (100 MHz, CDCl3) δ 170.5, 165.2, 143.5, 137.8, 136.9, 135.2, 129.4, 128.6, 127.8, 127.4, 126.8, 126.4, 61.6, 61.1, 56.1, 44.9, 41.7, 21.4, 14.08, 13.97; IR (film) νmax 2981, 1717, 1344, 1260, 1160, 1096 cm−1. HRMS-DART: calcd for C24H28NSO6 ([M + H]+) m/z 458.1632; found 458.1628.

General experimental procedure for the allene–imine [4+2] annulation (with the allene 11)

The allenoate (0.2 mmol) was added in one portion to a mixture of N-tosylimine (0.1 mmol), the aminophosphine catalyst (0.03 mmol), and 4-Å MS (10 mg) in CH2Cl2 (2 mL). The flask was capped with a Teflon cap under a N2 flow. The mixture was stirred at room temperature for 3 days. The resulting solution was concentrated and mesitylene was added as an internal standard to determine the yield and conversion. This mixture was dissolved in THF and then 1 N HCl was added to hydrolyze the imine. The aqueous phase was exacted with CH2Cl2. The combine extracts were washed with brine, dried (Na2SO4), and concentrated. The residue was purified through PTLC (hexane/EtOAc, 5:1) to give the product for ee determination.