Abstract
A short, high-yielding protocol involving the enantioselective α-chlorination of aldehydes has been developed for the enantioselective synthesis of C2-functionalized aziridines and N-alkyl terminal azetidines from a common intermediate. This methodology allows for the rapid preparation of functionalized aziridines in 50–73% overall yields and 88–94% ee, and azetidines in 22–32% overall yields and 84–92% ee. Moreover, we developed a scalable and cost-effective route to the key organocatalyst (54% overall yield, >95% dr).
Keywords: organocatalysis, azetdine, aziridine, enantioselective
Graphical Abstract
Azetidines comprise an important class of nitrogen-containing heterocycles due to both their biological significance and increasing use in medicinal chemistry. 1 While recent approaches have made progress in the ability to access various types of azetidines in an enantioselective fashion, 2-alkyl substituted azetidines are notably scarce in the literature.2,3
Similarly, aziridines represent an important class of nitrogen-containing heterocycles that have garnered significant interest from the synthetic chemistry and chemical biology communities over the past two decades.4 Aziridine’s intrinsic ring strain positions them as synhetically valuable intermediates capable of undergoing highly regio- and stereoselective ring opening transformations.5 Additionally, the aziridine moiety is found in compounds that possess interesting biological properties including antitumor or antibiotic activity. Together, these properties have shown aziridines to be useful synthetic intermediates and attractive targets in medicinal chemistry and the total synthesis of natural products.6,7
We recently reported a three step, one-pot protocol involving the enantioselective α-chlorination of aldehydes, subsequent reductive amination with a primary amine, followed by SN2 displacement to afford chiral N-alkyl terminal aziridnes.8 Under this protocol, yields and enantioselectivities in many cases were moderate and variable due to epimerization of the α-chlorinated aldehyde 3 during reductive amination. We sought to address these issues as well as provide a facile route to the synthesis of 2-substituted azetidines through a common, configurationallystable intermediate.
According to the seminal work of Jorgensen and coworkers for the organocatalytic α-chlorination of aldehydes, in situ reduction of the α-chloro aldehydes with NaBH4 to the corresponding β-chloro alcohols 5 would occur without loss of enantioselectivity.9 We envisioned that these configurationallystable α-chloro alcohols 5 could function as bifunctional chiral building blocks.
Essentially, we reasoned that functionalization of this common intermediate with either an amine (Route A) or nitrile group (Route B) followed by an intramolecular cyclization strategy could provide access to both aziridines and azetidines (Figure 1). As disclosed in our previous communication, catalyst 2 was optimal in terms of enantioselectivity and compatibility for the purposes of accessing the desired chloro alcohol.8,10,11 This catalyst was difficult to obtain commercially. In our hands, we found that literature procedures to prepare 2 resulted in variable yields and an inseparable mixture of our desired catalyst with the meso isomer.12,13 Recrystallization of the final product was achieved, successfully increasing the purity of the desired isomer to >92%. However, these conditions limited the scale and required careful temperature regulation during recrystallization. Therefore, we sought to improve the synthesis to both increase the yield and reliability of the protocol, with the goal of being able to rapidly obtain the pyrrolidine catalyst on multi-gram scale.
Figure 1.
Top: First-generation organocatalytic approach for the enantioselective synthesis of N-alkyl aziridines. Bottom: Envisioned second generation organocatalytic approach for the enantioselective synthesis of C2-functionalized aziridines and azetidines.
We successfully optimized the synthetic route as outlined in Scheme 1. Notably, we found that flash column chromatography of the penultimate intermediate 10 with a 7:3 hexanes/toluene solvent system provided excellent separation of the desired isomer. Additionally, increasing the amount of Wilkinson’s catalyst in the allyl deprotection to 10 mol% increased the yield of the deprotection to 71%. These improvements allowed for the gram-scale production of catalyst 2 in >54% overall yield and <95% dr.
Scheme 1.
Scalable, cost-effective synthesis of organocatalyst 2.
With sufficient organocatalyst for α-chlorination in hand, our attention was turned towards the synthesis of the aziridines. Preparation of β-chloro alcohols 5 proceeded via facile asymmetric organocatalytic chlorination and in situ sodium borohydride reduction of achiral aldehydes 1. To transform primary alcohols into bifunctional electrophiles, the β-chloro alcohols 5 were treated with triflic anhydride and lutidine in CH2Cl2 at 0°C. These conditions smoothly generated the desired triflates, which were then immediately treated with benzylamine to generate β-chloro amines in good yields for this one-pot procedure. The resulting amines were of sufficient purity to subject directly to intramolecular cyclization. Potassium hydroxide in THF/H2O at 65 °C promoted clean cyclization to afford the desired chiral N-alkyl terminal aziridines 11–15 in 50–73% overall yield and 88–94% ee from the readily available aldehyde starting materials. This mild, two-pot protocol from the common configurationally stable 2-chloro alcohol intermediate represents a robust approach to access aziridines, and this optimized approach avoids epimerization of the α-chloro aldehyde and extremely low temperature reaction conditions necessary in previous methodologies.8
We next focused on the synthesis of the azetidines (Figure 1, route B), as there are very few approaches for their preparation. Starting from the common 2-chloro alcohol intermediate, we planned to access the β-chloro nitrile intermediates 17–21 through analogous activation of the primary alcohol as the triflate, followed by displacement with potassium cyanide. Initial conditions using potassium cyanide in CH2Cl2, THF, or acetonitrile led to poor conversion at room temperature and elimination at elevated temperature. However, we found that potassium cyanide in acetonitrile, with 18-crown-6 as an additive, was effective in conversion to the desired nitrile after 24 hours at room temperature. Under these optimized conditions, displacement of the in situ prepared triflate with potassium cyanide facilitated formation to the desired β-chloro nitrile 17–21 in 66–86% yield. While these products were relatively stable at room temperature in CH2Cl2 solution, elevated temperatures resulted in the formation of an elimination byproduct. Similarly, the concentrated products were prone to elimination, and decomposed within hours at room temperature.
Related to the instability of the β-chloro nitriles, we found that conventional methods to reduce the nitrile such as lithium aluminum hydride and hydrogenation led to either elimination or dehalogenation of the starting material. After screening various mild nitrile reduction conditions we found that using an indium (III) chloride-sodium borohydride system to be generally applicable across substrates in excellent yield and sufficient purity to carry forward without purification towards cyclization.
With the γ-chloro amine 22 in hand, we focused on the cyclization conditions to afford the desired azetidine 6. We initially attempted the same conditions that were successful in facilitating the 3-exo-tet cyclization of the β-chloro amine to aziridines. However, treatment with KOH in THF/H2O at 65 °C led to minimal consumption of the starting material. Additionally, a variety of modifications to the base and solvent system promoted elimination to an undesired byproduct 30. To find conditions that might promote 4-exo-tet cyclization, we performed a screen of a broad selection of organic and inorganic bases, solvents, and temperatures (Table 1).
Table 1.
Conditions for the base-induced cyclization of β-chloro amines to azetidines 29 and elimination product 30.
| |||||
|---|---|---|---|---|---|
| Entry | Base | Solvent | Temperature | Additive | Conversion (29:30) |
| 1 | K2CO3 | NMP | 25 °C | -- | 0% |
| 2 | K2CO3 | NMP | 120 °Cb | -- | 40% (2:1) |
| 3 | K2CO3 | NMP | 180 °Cb | -- | 80% (2:1) |
| 4 | -- | THF:H2Oa | 150 °Cb | -- | 75% (1.5:1) |
| 5 | NaH | DMF | 25 °C | -- | 0% |
| 6 | NaH | DMF | 65 °C | -- | 0% |
| 7 | NaH | DMF | 25 °C | 15-C-5 | trace (1:2) |
| 8 | NaH | DMF | 65 °C | 15-C-5 | <5% (0:1) |
| 9 | LHMDS | DMF | 25 °C | -- | 60% (0:1) |
| 10 | LHMDS | DMF | 25 °C | -- | 100% (0:1) |
| 11 | K2CO3 | DMF | 25 °C | -- | 0% |
| 12 | K2CO3 | DMF | 120 °Cb | -- | 10% (2:1) |
| 13 | K2CO3 | DMF | 25 °C | AgNO3 | 0% |
| 14 | K2CO3 | DMF | 65 °C | AgNO3 | 0% |
| 15 | KOH | THF:H2Oa | 65 °C | -- | <5% (3:1) |
| 16 | KOH | THF:H2Oa | 120 °Cb | -- | 20% (3:1) |
| 17 | KOH | THF:H2Oa | 170 °Cb | -- | 100% (3:1) |
All reactions run at 0.1 mmol, 0.125M in solvent. Conversion determined by LCMS and NMR
This screen revealed that many conditions which promoted cyclization also facilitated elimination to the olefin 30. We identified optimal conditions which provided the desired azetidine in a 3:1 ratio with the competing elimination pathway. Specifically, potassium hydroxide in THF/H2O (1:1), under high thermal conditions (170 °C) was effective in conversion to the desired azetidine 29. Microwave irradiation for 1 hour at this temperature resulted in full consumption of the γ-chloro amine starting material.
Of note, the major by-product was identified as the olefin resulting from chloride elimination. γ-Chloro amines with branched alkyl groups or aryl groups β to the secondary chloride were especially susceptible to elimination, as reflected in the decreased yields for 34. This can be rationalized by the high energetic requirements necessary to enable ring closure to the highly strained 4-member azetidine ring, which rapidly promoted elimination to the conjugated olefin. Gratifyingly, the azetidine and olefin products were readily seperated by flash column chromatography.
In order to determine the scope of this methodology, we prepared several 2-alkyl azetidines from readily available aldehydes. Nitrile reduction and azetidine cyclization were performed sequentially without purification of the γ-chloroamine in all cases. This provided the desired azetidines in 44–55% yield. Attempts to induce cyclization to 34 resulted in nearly complete elimination to the conjugated byproduct, with less than 5% recovered product. Importantly, this approach provides the azetidines without N-functionalization, allowing for rapid derivitization using robust synthetic methods such as reductive amination and N-alkylation.
In summary, we have developed an optimized three-step procedure for the enantioselective synthesis of N-alkyl terminal aziridines and azetidines with alkyl substituents at the C2 position of each heterocycle. This methodology allows for the rapid preparation of the functionalized aziridines in 50–73% overall yields and 88–94% ee, and functionalized azetidines in 22–32% overall yields and 84–92% ee. This new method addresses deficiencies in our first generation approach for the synthesis of aziridines while facilitating the synthesis of azetidines through a common bifunctional intermediate. Alternative methods to access azetidines that do not rely on the chiral pool is a demonstrated need in natural products synthesis and medicinal chemistry applications, and the ability to employ simple aldehydes and organocatalysts towards their synthesis allows for straightforward access to either enantiomer. Additional refinements and applications of this methodology to the synthesis of biologically relevant small molecules are under development and will be reported in due course.
Scheme 2.
Protocol for enantioselective synthesis of N-alkyl terminal aziridines.14
Scheme 3.
Protocol for the synthesis chiral β-chloro nitriles 17–21.
Scheme 4.
Mild reduction of β-chloro nitriles to γ-chloro amines 23–27.
Scheme 5.
Protocol for intramolecular cyclization to C2-functionalized azetidines 31–34.15
Acknowledgments
Financial support from the Warren Family and Foundation is gratefully acknowledged, and both T.J.S. and M.C.O. acknowledge partial support from the ACS MEDI division and ACS MEDI Predoctoral Fellowships.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References and notes
- 1.Bott TM, West FG. Heterocycles. 2012;84:223. [Google Scholar]
- 2.Brandi A, Cicchi S, Cordero FM. Chem Rev. 2008;108:3988. doi: 10.1021/cr800325e. [DOI] [PubMed] [Google Scholar]
- 3.Wells JN, Tarwater OR. J Pharm Sci. 1971;60:156. doi: 10.1002/jps.2600600140. [DOI] [PubMed] [Google Scholar]
- 4.Yudin AK. Aziridines and Epoxides in Organic Chemistry. 2006 [Google Scholar]
- 5.Müller P, Fruit C. Chem Rev. 2003;103:2905. doi: 10.1021/cr020043t. [DOI] [PubMed] [Google Scholar]
- 6.Degennaro L, Trinchera P, Luisi R. Chem Rev. 2014 doi: 10.1021/cr400553c. ASAP. [DOI] [PubMed] [Google Scholar]
- 7.Majumdar KC, Chattopadhyay SK. Heterocycles in Natural Product Synthesis. Wiley-VCH Verlag GmbH & Co. KGaA; Weinheim, Germany: 2011. [Google Scholar]
- 8.Fadeyi OO, Schulte ML, Lindsley CW. Org Lett. 2010;12:3276. doi: 10.1021/ol101276x. [DOI] [PubMed] [Google Scholar]
- 9.Halland N, Braunton A, Bachmann S, Marigo M, Jørgensen KA. J Am Chem Soc. 2004;126:4790. doi: 10.1021/ja049231m. [DOI] [PubMed] [Google Scholar]
- 10.O’Reilly MC, Lindsley CW. Org Lett. 2012;14:2910. doi: 10.1021/ol301203z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.O’Reilly MC, Lindsley CW. Tetrahedron Lett. 2012;53:1539. doi: 10.1016/j.tetlet.2011.12.063. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Aldous DJ, Dutton WM, Steel PG. Tetrahedron: Asymmetry. 2000;11:2455. [Google Scholar]
- 13.Kemppainen EK, Sahoo G, Valkonen A, Pihko PM. Org Lett. 2012;14:1086. doi: 10.1021/ol203486p. [DOI] [PubMed] [Google Scholar]
- 14.Representative experimental for the synthesis of N-alkyl terminal aziridine. (R)-1-benzyl-2-octylaziridine (13). To a round bottom flask equipped with a stir bar, 4-(5,5-dimethyl-1,3-dioxan-2-yl)butanal (0.25 g, 1.34 mmol) was dissolved in CH2Cl2 (3.84 mL) at 0°C and to the stirring solution was added (2R,5R)-2,5-Diphenylpyrrolidine (0.03 g, 0.13 mmol) and then N-chlorosuccinimide (0.233 g, 1.75 mmol). Solution was maintained at 0 °C, and stirring was continued for 12 hours, at which point starting material was consumed by 1HNMR. The reaction was diluted with MeOH (3.84 mL) and cooled to 0°C, then NaBH4 (0.254 g, 6.72 mmol) was slowly added while stirring. The reaction was stirred for additional 30 min. at the same temperature, then quenched with H2O and extracted three times with EtOAc. The organic fractions were combined, washed with brine, and dried over Na2SO4. The organic layer was concentrated in vacuo and a crude oil was purified by column chromatography (4:1 Hex/EtOAc) to yield desired product as a clear oil in 0.23 g (76%). 1HNMR (400.2 MHz, CDCl3) δ (ppm): 4.49 (t, J= 4.4 Hz, 1H), 4.08 (m, 1H), 3.81 (dd, J= 12.2, 4.0 Hz, 1H), 3.69 (dd, J= 11.7, 7.0 Hz, 1H), 3.62 (d, J= 10.8 Hz, 2H), 3.44 (d, J= 10.8 Hz, 2H), 2.05 (s, 1H), 2.00–1.74 (m, 4H), 1.20 (s, 3H), 0.74 (s, 3H); 13C NMR (100.6 MHz, CDCl3) δ (ppm): 101.2, 76.6, 66.8, 64.7, 31.2, 30.1, 28.3, 22.9, 21.7; HRMS (TOF, ES+) C10H19ClO3 [M+H]+ calc’d 223.1021, found 223.1020; Specific rotation [α]23/D = −3.0° (c = 1.0, CH3Cl). To a round bottom flask equipped with a stir bar, alcohol (0.10 g, 0.45 mmol) and 2,6-lutidine (0.241 g, 2.25 mmol) were dissolved in CH2Cl2 (2.25 mL) and cooled to 0°C. Tf2O (0.54 mL, 1M in CH2Cl2, 0.54 mmol) was then added dropwise, and the solution was left to stir at 0°C for 30 minutes. At this time, the triflate solution was taken up by syringe and added dropwise to a stirring solution of benzylamine (0.48 g, 4.50 mmol) in CH2Cl2 (1.00 mL) at 0°C. The flask containing the triflate solution was washed 2 times with CH2Cl2 (1.00 mL) and added to reaction. The solution was then slowly warmed to ambient temperature and stirred overnight. The reaction was quenched with NaHCO3, diluted with CH2Cl2, and extracted 3 times with CH2Cl2. The organic layers were combined and washed with brine. The organic layer was dried over Na2SO4 and concentrated in vacuo to yield the crude product. The crude product was isolated as a pale yellow oil quantitatively and carried forward directly to the next step. In an open microwave vial equipped with a stir bar, crude amine (0.14 g, 0.45 mmol) was dissolved in 1:1 THF/H2O (2.99 mL), followed by the addition of KOH (0.16 g, 2.92 mmol). The microwave vial was sealed and stirred overnight at 65 °C. The reaction mixture was allowed to cool to ambient temperature, then extracted 3 times with EtOAc. The organic layers were combined and dried over Na2SO4, then concentrated in vacuo to give the crude product a pale yellow oil. Purification by flash column chromatography (1:1 Hex/EtOAc) afforded the desired product as pale yellow oil in 0.11 g, (85%). 1HNMR (400.2 MHz, CDCl3) δ (ppm): 7.23 (m, 5H), 4.28 (t, J= 5.0 Hz, 1H), 3.50 (d, J= 10.8 Hz, 2H), 3.43 (d, J= 13.6 Hz, 1H), 3.27 (m, 3H), 1.61 (m, 2H), 1.55 (d, J= 3.4 Hz, 1H), 1.47 (m, 3H), 1.32 (d, J= 6.0 Hz, 1H), 1.10 (s, 3H), 0.63 (s, 3H); 13C NMR (100.6 MHz, CDCl3) δ (ppm): 139.5, 128.4, 128.3, 127.0, 101.7, 77.3, 65.1, 39.5, 34.0, 32.6, 30.2, 27.6, 23.1, 21.9; HRMS (TOF, ES+) C17H25NO2 [M+H]+ calc’d 276.1886, found 276.1886; Specific rotation [α]23/D = +4.0° (c = 1.0, CH3Cl).
- 15.Representative experimental for the synthesis of C2-functionalized azetidine. (S)-2-(3-fluoro-4-methylbenzyl)azetidine (32). To a round bottom flask equipped with a stir bar, alcohol 39 (0.19g, 0.89 mmol) and 2,6-lutidine (0.48g, 4.44 mmol) were dissolved in CH2Cl2 (4.4 mL) and cooled to 0°C. Tf2O (0.30g, 1.07 mmol) was then added dropwise, and solution left to stir at 0°C for 30 minutes. Reaction was diluted with CH2Cl2 and washed with H2O. The organic fraction was washed with brine, dried over Na2SO4, and concentrated in vacuo. The crude oil was carried on directly to the next step. The crude triflate was dissolved in MeCN (4.4 mL), then 18-C-6 (0.05g, 0.18 mmol) was added to the mixture and allowed to dissolve. KCN (0.578g, 8.88 mmol) was then added, and the reaction was allowed to stir for 12 h at ambient temperature. The reaction was then quenched with NaHCO3 and extracted three times with CH2Cl2. The organic fractions were combined and dried over Na2SO4. The organic layer was concentrated in vacuo, and purification by flash chromatography (4:1 Hex/EtOAc) afforded the desired product as a clear/rose colored oil in 0.11 g (70%). 1HNMR (400.2 MHz, CDCl3) δ (ppm): 4.45 (dd, J= 10.4, 1.6 Hz, 1H, (tt, J= 10.2, 3.5 Hz, 1H), 3.60 (m, 2H), 3.38 (d, J= 8.9 Hz, 1H), 3.20 (m, 3H), 2.54 (m, 1H), 2.07 (m, 2H), 1.71 (dt, J= 15.4, 5.0 Hz, 1H), 0.85 (s, 3H), 0.82 (s, 3H); 13C NMR (100.6 MHz, CDCl3) δ (ppm): 117.9, 75.9, 75.7, 72.7, 68.5, 53.86, 36.7, 35.4, 28.2, 23.2, 23.1; HRMS (TOF, ES+) C11H18ClNO2 [M+H]+ calc’d 232.1025, found 232.1025; Specific rotation [α]23/D = +6.4° (c = 1.1, CH3Cl). To a microwave vial with magnetic stir bar was added InCl3 (0.10 g, 0.43 mmol) and NaBH4 (0.05 g, 1.30 mmol). Anhydrous THF (1.0 mL) was added and the heterogeneous mixture was stirred under argon for 1h. 25 (0.10 g, 0.43 mmol) was then slowly added as a solution in THF (1.0 mL). Reaction mixture was allowed to stir at ambient temperature for 4 h. The reaction was quenched by dropwise addition (CAUTION: results in rapid generation of gas) of H2O (1.0 mL) and the solution was heated to 75 °C for 30 min. MeOH (2.0 mL) was then slowly added, and stirring continued at 75 °C for 30 min. The reaction mixture was allowed to cool to room temperature, then filtered through a celite pad to remove solid indium byproducts. The resulting colorless solution was concentrated in vacuo to provide 32 in 96% yield. The crude residue was carried forward directly to the next step. To an open microwave vial equipped with a stir bar, the crude residue from the previous step was dissolved in a solution of 1:1 THF/H2O (3.40) and KOH (0.16 g, 2.76 mmol) was then added. The vial was sealed and submitted to microwave irradiation at 170°C. for 1 h. The biphasic solution was extracted three times with EtOAc and washed with brine. The organic fractions combined, dried over Na2SO4, and concentrated in vacuo. Purification by flash column chromatography (9:0.9:0.1 CH2Cl2/MeOH/NH4OH) afforded desired product as a light yellow oil in 45 mg (55%). 1HNMR (400.2 MHz, MeOD) δ (ppm): 3.89 (dd, J= 12.5, 3.6 Hz, 1H), 3.63 (m, 1H), 3.62 (d, J= 12.0 Hz, 1H), 3.57 (d, J= 12.0 Hz, 1H), 3.35 (m, 3H), 1.97 (m, 1H), 1.79 (d, J= 7.0 Hz, 1H), 1.50 (m, 4H), 1.35 (d, J= 3.4 Hz, 1H), 0.94 (s, 3H), 0.88 (s, 3H); 13C NMR (100.6 MHz, MeOD) δ (ppm): 81.8, 81.0, 79.3, 77.2, 38.1, 29.6, 29.3, 29.1, 24.1, 23.0, 22.6; specific rotation [α]D 20 = +1.7 (c = 1.0, CHCl3); HRMS (TOF, ES+) C11H22NO3S [M+H]+ calc’d 200.1574, found 200.1575