A telescopic one-pot synthesis of β-lactam rings using amines as a convenient source of imines

Abstract

A facile synthetic approach to substituted β-lactams was designed, using secondary benzylic amines and acid chlorides as starting materials. The reactions proceeded smoothly and all the products were obtained in good yields.

---

Since the introduction of penicillin in the 1940s,¹ antibiotics have become a milestone of modern medicine.² Penicillins, cephalosporins and monobactams all belong to the family of β-lactam antibiotics,³ in which the β-lactam ring is part of the core structure.⁴ Antibiotics now face a critical and unprecedented challenge resulting from the combination of two different problems. First, antibiotic resistance to commonly used drugs is increasing at an alarming rate.⁵ Second, the development of new antibiotics has become so expensive and unattractive for pharmaceutical companies that the antibiotic pipeline is running dry.⁶ The need to overcome the problem caused by drug-resistant bacteria⁷ has motivated chemists to modify the structure of β-lactam antibiotics, aiming to render it insensitive to the β-lactamase attack.⁸

Today, despite an enormous number of synthetic strategies to prepare the β-lactam ring,⁹ the Staudinger approach¹⁰ using ketenes and imines is regarded as the most effective and versatile procedure for the synthesis of 2-azetidinone rings (Scheme 1).¹¹ However, the preparation of labile imines¹² requires an additional synthetic step, and several issues affect the yield due to the existence of equilibrium conditions between the reactants and the corresponding imines. Taking these factors into consideration, we planned to generate imines by oxidation of the corresponding amines in order to develop an effective procedure to facilitate the access to substituted β-lactam rings.¹³

Scheme 1 A modified Staudinger β-lactam synthesis using chloramines.

Following our previous work on both β-lactams¹⁴ and chloramines,¹⁵ we designed a modified Staudinger β-lactam synthesis using N-chloramines as a convenient source of imines (Scheme 1). The in situ-generated imine may subsequently react with a preformed ketene via a [2 + 2] cycloaddition to afford the corresponding β-lactam.

The first step was the formation of a chloramine, followed by subsequent elimination with base to form the corresponding imine. N-Benzylamine 1 (1.0 mmol), chosen as a model substrate, was dissolved in dry Et₂O and treated with N-chlorosuccinimide (NCS, 1.1 mmol) at room temperature in an inert atmosphere. After 1 h, TLC analysis showed complete consumption of the starting amine, and the aqueous work-up of the reaction mixture afforded chloramine 2 in an almost quantitative yield (Scheme 2).¹⁶

Scheme 2 Preparation of imines directly from secondary amines.

In the following step, the resulting chloramine 2 (1 mmol) dissolved in Et₂O was treated overnight at room temperature with an ethanolic solution of KOH (1.1 mmol), to afford the corresponding imine 3 in good yield (71%, Scheme 2).

A set of bases in different solvents was then screened to optimise the yield of imine 3 without isolating the intermediate chloramine 2.¹⁷ The best results were obtained by performing both the formation of chloramine and the elimination reaction to give the imine in sequence in acetonitrile, using triethylamine (1.5 mmol) as base (Table 1, entry 3). The imine formation was complete in just 3 hours at room temperature.

Table 1 Screening of bases and solvents for imine formation

Entry	Base	Solvent	Yield^a (%)
a Yields were determined by NMR from crude reaction mixture.b DBU = 1,8-diazabicyclo[5.4.0]undec-7-ene.c Anhydrous K2CO3.d DABCO = 1,4-diazabicyclo[2.2.2]octane.e TMG = 1,1,3,3-tetramethylguanidine.f DMAP = 4-dimethylaminopyridine.
1	DBU^b	CH2Cl2	49
2	DBU	CH3CN	54
3	NEt3	CH3CN	99
4	NEt3	CH2Cl2	85
5	NEt3	Et2O	81
6	K2CO3^c	CH3CN	8
7	DABCO^d	CH3CN	—
8	TMG^e	CH3CN	—
9	NEt3/DMAP^f	CH3CN	61

---

Having established a reliable and easy procedure to convert amines into imines, we then turned our attention to study the [2 + 2] cycloaddition reaction of in situ-generated imines with ketenes.

Initially, to the imine 3 was added a solution of phenoxyacetyl chloride that had been previously treated with an equimolar amount of triethylamine to generate the corresponding ketene. The pre-formed ketene reacted smoothly with the imine 3 giving a cis-β-lactam as the product (see Scheme in Table 2). In order to find the best route to prepare 2-azetidinone rings, we carried out several experiments in parallel, reacting amine 1 with various amounts of phenoxyacetyl chloride and triethylamine, as summarised in Table 2. The best results were achieved dissolving the phenoxyacetyl chloride in acetonitrile, and using 2 equivalents of acid chloride and triethylamine for each equivalent of N-methylbenzylamine (Table 2, entry 4). Once it was established that in situ imine and ketene formation, as well as the subsequent 2 + 2 rearrangement, proceeded smoothly in acetonitrile, we then explored the possibility of developing the synthesis of β-lactams 4 via a one-pot two-step protocol. Therefore, N-methylbenzylamine 1 (1 mmol) dissolved in dry acetonitrile (10 ml) was first chlorinated with NCS (1.1 mmol), treated with excess triethylamine (3.5 mmol) for 1 h under argon at rt, and finally reacted overnight with phenoxyacetyl chloride (2 mmol) in dry acetonitrile (10 ml).¹⁸ Under these optimised reaction conditions, the cis-β-lactam 4 was isolated (85%) as the major product.¹⁹

Table 2 Screening of solvents and reagent quantities for the final Staudinger β-lactam synthesis^a

Entry	Amine 1 (eq.)	PhOCH2COCl (eq.)	Solvent	Yield^b (%)
a Reactions were carried out at room temperature under Ar.b Isolated yields.
1	2.0	1.0	CH3CN	42
2	1.0	1.5	CH3CN	71
3	1.5	1.0	CH3CN	51
4	1.0	2.0	CH3CN	85
5	1.0	2.0	Toluene	16
6	1.0	2.0	CH2Cl2	69
7	1.5	2.0	THF	22

---

With these optimised reaction conditions, the scope and limitations of this protocol were studied applying it to various benzylic amines (Scheme 3, β-lactams 5–14). Benzylic amines with either an electron-withdrawing (2-Br, 3-CF₃) or an electron-donating group (3-MeO or 3-MeO, 4-MeO) on the aryl ring proved to be suitable substrates for this reaction, affording the corresponding β-lactams 5–8 in high yields. The position and activating or deactivating nature of substituents on the rings did not significantly affect the diastereoselectivity of the reaction (cis/trans > 9/1) and they also had negligible influence on the yield of the isolated products (ranging from 72% to 79%).

Scheme 3 Synthesis of α-phenyloxy-β-lactams with aryloxyacetyl chlorides and an array of benzyl amine derivatives.

Ethyl, isopropyl and benzyl groups on the benzylic nitrogen were well tolerated, providing the corresponding β-lactams 9–11 in high overall isolated yields. Conversely, the reaction with N-benzyl anilines (Scheme 3, β-lactam 12) was problematic and only traces of product were detected.²⁰

In the case of β-lactam 13, the product was formed in good yield as judged by ¹H NMR analysis of a sample of the crude reaction mixture. Unfortunately the presence of many side products rendered the reaction unviable and any attempt to purify the product proved unsuccessful. It is worth noting that aryloxyketenes possessing a substituent (Me, OMe, Cl, NO₂) on the aromatic ring (Scheme 3, β-lactams 14–17) were compatible with the designed method. Moreover, in general, acid chlorides containing electron-rich substituents on the phenyl ring provided better yields than those bearing electron-withdrawing groups (Scheme 3).

After these encouraging results from the initial experiments, the general scope of this protocol was further explored by reacting the in situ-generated imine 3 with a set of different acid chlorides used as precursors of ketenes. The reaction worked well also with methoxyacetyl chloride (Scheme 4, β-lactams 18–22). The presence of a longer side chain on the ketene did not significantly affect the final yield (Scheme 4, β-lactams 22). Although the yields of these 2-azetidinones were somewhat lower than with phenoxyacetyl chloride, they should be considered satisfactory, since reactions are performed in a one-pot, three-step fashion.

Scheme 4 Synthesis of β-lactams with methoxyacetyl chloride.

Surprisingly, when the phenoxyacetyl chloride was replaced with phenylacetyl chloride, no product was obtained when the reaction was carried out at room temperature. We reoptimised the reaction conditions and we were delighted to find that heating the reaction mixture overnight at 60 °C in the ketene formation step, the Staudinger reaction proceeded as smoothly as before, affording the β-lactam 23 in reasonable yields (Scheme 5).

Scheme 5 Synthesis of β-lactams with substituted acetyl chlorides.

The reaction yields decreased slightly for substituted phenylacetyl chlorides bearing electron-withdrawing groups on the aromatic ring (Scheme 5, compare compounds 23 and 24). In contrast, electron-donating groups increased the yields significantly, giving an overall quantitative yield of cis and trans isomers (Scheme 5, see β-lactams 25 and 26). We also attempted to carry out the reaction with aliphatic acyl chlorides, but the annulation reaction to give β-lactam 27 was unsuccessful (Scheme 5).²¹

It is important to note that the reaction with (phenylthio)acetyl chloride failed to give the desired β-lactam 28 at room temperature, but with the procedure used for phenylacetyl chlorides the reaction proceeded smoothly and in good yield, giving the trans isomer as the only isolated product.

Next, we decided to explore the scope of this modified Staudinger procedure combining the ketene derived from the phenylacetyl chloride with a set of imines generated from substituted benzylic amines. Generally, substituents on the aryl group of N-methylbenzylamine do not seem to influence the outcome of the reaction (Scheme 5, β-lactams 29 and 30). When N-methylfurfurylamine was employed, product 31 was identified in the crude NMR, but as for β-lactam 13, isolation and purification proved very difficult and the product was not isolated in acceptable purity.

The reaction failed to provide the desired β-lactam using aliphatic amines with either of the protocols. The reaction was tested with N-methylbutylamine and phenoxyacetyl chloride, varying the base²² used in imine formation as well as solvent and temperature of the reaction (Scheme 6).

Scheme 6 Synthesis of β-lactams with butylamine and phenoxyacetyl chloride.

While the synthesis of the chloramine proceeded smoothly, both imine formation and the subsequent Staudinger reaction with ketene to give a β-lactam resulted in a complex mixture of products, unidentifiable by NMR.

Despite some unsuccessful attempts, we were pleasantly surprised by the high number of unknown compounds (>85%)²³ prepared using this procedure, suggesting that our methodology may provide a direct and alternative route to β-lactam structures.

On the basis of above results and previously published mechanisms,²⁴ we postulate the most plausible reaction pathway for the overall conversion of secondary benzyl amines and acid chlorides into β-lactams (Scheme 7). In the first step, NCS promotes the chlorination of the N-alkylamine 32, releasing an equivalent amount of succinimmide. In the second step, NEt₃ abstracts a proton from both chloramine 33 and acid chloride 34 generating the imine 35 and ketene 36 respectively. In the following step, the imine nitrogen attacks the ketene giving a zwitterionic intermediate, 37. The subsequent intramolecular ring closure of the zwitterionic intermediate delivers β-lactam 38. The cis or trans configuration of the final β-lactam is always the result of a complex dynamic balance between two competitive reactions involving the conrotatory ring closure on one hand and the imine isomerization (39) of the zwitterionic intermediate 37 on the other.

Scheme 7 A plausible reaction pathway for the overall conversion of secondary benzyl amines and acid chlorides into β-lactams.

In general, electron-donating (EDG) ketene substituents accelerate direct ring closure, promoting the formation of cis-β-lactam 38. On the contrary, weak electron-donating and electron-withdrawing (EWG) ketene substituents slow down direct ring closure, strongly favouring trans-β-lactam isomer.

Conclusions

A facile synthetic approach to substituted β-lactams was designed, starting from secondary benzyl amines and acid chlorides. Instead of classical imines, which are labile molecules, secondary amines are used expanding the molecular diversity of the traditional Staudinger methodology. Generally, the reactions proceeded smoothly and all synthesised products were isolated in good to high yields. This modification of the traditional Staudinger synthesis, where both imine and ketene are formed in situ from cheap starting materials and no additives apart from a base are used, reduces both the costs and the purification steps of the procedure to a minimum.

Acknowledgements

Financial support from the Università degli Studi di Cagliari (PRID 2015) and the Regione Autonoma della Sardegna (Project L.R. 7/2007, annualità 2012, CUP F71J120001090002) is gratefully acknowledged. We are also grateful to Fondazione Banco di Sardegna (Project: Prot. U875.2014/AI.758.MGB, Prat.2014.0803).

Notes and references

1. American Chemical Society International Historic Chemical Landmarks. Discovery and Development of Penicillin. Retrieved from http://www.acs.org/content/acs/en/education/whatischemistry/landmarks/flemingpenicillin.html.
2. Antibiotics have not only saved millions of human lives, but they have opened up significant new vistas for research that would have been unmanageable without them. L. Demain and R. P. ElanderAntonie van Leeuwenhoek 1999, 75 5–19; W. Dürckheimer, J. Blumbach, R. Lattrell and K. H. Scheunemann, Angew. Chem., Int. Ed. Engl., 1985, 24, 180–202; A. Cometta and M. P. Glauser, Schweiz. Med. Wochenschr., 1988, 118, 1817–1822; R. Bud, Penicillin: Triumph and Tragedy, Oxford University Press, Oxford, 2009; E. Torok, E. Moran and F. Cooke, Oxford Handbook of Infectious Diseases and Microbiology, Oxford University Press, Oxford, 2009.
3. “List of β-lactam antibiotics”, retrieved from https://en.wikipedia.org/wiki/β-lactam_antibiotic.
4. P. G. Sammes, Chem. Rev., 1976, 76, 113–155; R. B. Morin and M. Gorman, Chemistry and Biology of β-Lactam Antibiotics, Academic Press, New York, 1982, vol. 1–3; W. Dürkheimer, J. Blumbach, R. Lattrell and K. H. Scheunemann, Angew. Chem., 1985, 97, 183–205; W. Dürkheimer, J. Blumbach, R. Lattrell and K. H. Scheunemann, Angew. Chem., Int. Ed. Engl., 1985, 24, 180–202; Recent Progress in the Chemical Synthesis of Antibiotics and Related Microbial Products, ed. R. Southgate, C. Branch, S. Coulton, E. Hunt and G. Lukacs, Springer-Verlag, Berlin, 1993, vol. 2, p. 621; R. Southgate, M. Jayaraman, A. R. A. S. Deshmukh and B. M. Bhawal, Contemp. Org. Synth., 1994, 1, 417–432; E. L. Setti, D. Davis, T. Chung and J. McCarter, Bioorg. Med. Chem. Lett., 2003, 13, 2051–2053; B. K. Banik, β-Lactams: Unique Structures of Distinction for Novel Molecules, Springer, 2013.
5. D. M. Shlaes, Antibiotics: The Perfect Storm, Springer, Dordrecht: Heidelberg, London, NY, USA, 2010; J. N. Pendleton, S. P. Gorman and B. F. Gilmore, Expert Rev. Anti-Infect. Ther., 2013, 11, 297–308; W. Qin, M. Panunzio and S. Biondi, Antibiotics, 2014, 3, 193–215.
6. This has led to an alarming situation where some bacterial infections are difficult, if not impossible, to cure. Over time this problem looks set to get worse. D. I. Andersson and D. Hughes, Nat. Rev. Microbiol., 2010, 8, 260–271; D. Jabes, Curr. Opin. Microbiol., 2011, 14, 564–569; T. R. Frieden, Antibiotic resistance threats in the United States, Cdc, 2013, pp. 22–50; T. R. Frieden, N. Engl. J. Med., 2015, 373, 1748–1754.
7. M. G. P. Page, Drug Resist. Updates, 2000, 3, 109–125; S. Driussi, Antibiotic Resistant Bacteria, Hayle Medical, 2015.
8. S. M. Drawz and R. A. Bonomo, Clin. Microbiol. Rev., 2010, 23, 160–201; D. M. Shlaes, Ann. N. Y. Acad. Sci., 2013, 1277, 105–114; R. R. Watkins, K. M. Papp-Wallace, S. M. Drawz and R. A. Bonomo, Front. Microbiol., 2013, 4, 1–8; S. M. Drawz, K. M. Papp-Wallace and R. A. Bonomo, Antimicrob. Agents Chemother., 2014, 58, 1835–1846; I. Olsen, Eur. J. Clin. Microbiol. Infect. Dis., 2015, 34, 1303–1308.
9. G. S. Singh, Tetrahedron, 2003, 59, 7631–7649; Heterocyclic Scaffolds I: β-Lactams, ed. B. K.Banik, Springer, Heidelberg, 2009, vol. 1; G. S. Singh and S. Sudheesh, ARKIVOC, 2014, 337–385.
10. H. Staudinger, Justus Liebigs Ann. Chem., 1907, 356, 51–123; W.-B. Wang and E. J. Roskamp, J. Am. Chem. Soc., 1993, 115, 9417–9420; G. Martelli, G. Spunta and M. Pannunzio, Tetrahedron Lett., 1998, 39, 6257–6260; M. Pannunzio, S. Bacchi, E. Campana, L. Fiume and P. Vicennati, Tetrahedron Lett., 1999, 40, 8495–8498; C. M. L. Delpiccolo, M. A. Fraga and E. G. Mata, J. Comb. Chem., 2003, 5, 208–210; A. Weatherwax, C. J. Abraham and T. Lectka, Org. Lett., 2005, 7, 3461–3463; F. L. Lin, H. M. Hoyt, H. Van Halbeek, R. G. Bergman and C. R. Bertozzi, J. Am. Chem. Soc., 2005, 127, 2686–2695; Y. Wang, Y. Liang, L. Jiao, D. M. Du and J. Xu, J. Org. Chem., 2006, 71, 6983–6990; L. Jiao, Y. Liang and J. Xu, J. Am. Chem. Soc., 2006, 128, 6060–6069; B. Li, Y. Wang, D.-M. Du and J. Xu, J. Org. Chem., 2007, 72, 990–997; Y. Liang, L. Jiao, S. Zhang, Z.-X. Yu and J. Xu, J. Am. Chem. Soc., 2009, 131, 1542–1549.
11. C. Palomo and J. M. Aizpurua, Chem. Heterocycl. Compd., 1999, 34, 1222–1236; C. Palomo, J. M. Aizpurua, I. Ganboa and M. Oiarbide, Eur. J. Org. Chem., 1999, 3223–3235; A. Arrieta, F. P. Cossio and B. Lecea, J. Org. Chem., 2000, 65, 8458–8464; A. E. Taggi, A. M. Hafez, H. Wack, B. Young, D. Ferraris and T. Lectka, J. Am. Chem. Soc., 2002, 124, 6626–6635; M. Köhn and R. Breinbauer, Angew. Chem., Int. Ed., 2004, 43, 3106–3116; S. Eguchi, ARKIVOC, 2005, 98–119; F. L. Lin, H. M. Hoyt, H. Van Halbeek, R. G. Bergman and C. R. Bertozzi, J. Am. Chem. Soc., 2005, 127, 2686–2695; L. Jiao, Y. Liang and J. Xu, J. Am. Chem. Soc., 2006, 128, 6060–6069; N. Fu and T. T. Tidwell, Tetrahedron, 2008, 64, 10465–10496; F. P. Cossío, A. Arrieta and M. A. Sierra, Acc. Chem. Res., 2008, 41, 925–936; J. Burés, M. Martín, F. Urpí and J. Vilarrasa, J. Org. Chem., 2009, 74, 2203–2206; Name reactions for carbocyclic ring formations, ed. J. J.Li, Wiley, Hoboken, N. J., 2010, p. 47; M. Zarei and A. Jarrahpour, Synlett, 2011, 2572–2576; Z. Yang and J. Xu, Tetrahedron Lett., 2012, 53, 786–789; I. Kim, S. W. Roh, D. G. Lee and C. Lee, Org. Lett., 2014, 16, 2482–2485; C. R. Pitts and T. Lectka, Chem. Rev., 2014, 114, 7930–7953; Z. Yang, N. Chen and J. Xu, J. Org. Chem., 2015, 80, 2–11.
12. A. T. Williamson, J. Am. Chem. Soc., 1934, 56, 2216–2218; G. Friedrichs and H. G. Wagner, Z. Phys. Chem., 2001, 215, 1601–1623; T. T. Tidwell, Eur. J. Org. Chem., 2006, 563–576; A. D. Allen and T. T. Tidwell, Eur. J. Org. Chem., 2012, 1081–1096; W. Qin, S. Long, M. Panunzio and S. Biondi, Molecules, 2013, 18, 12264–12289.
13. We took inspiration from nature, where aldehydes and ketones are generated in situ (via imines) by oxidation of corresponding amino groups.
14. D. Donati, C. Morelli, A. Porcheddu and M. Taddei, J. Org. Chem., 2004, 69, 9316–9318.
15. L. De Luca and G. Giacomelli, Synlett, 2004, 2180–2184; L. De Luca, G. Giacomelli and G. Nieddu, Synlett, 2005, 223–226; R. Cadoni, A. Porcheddu, G. Giacomelli and L. De Luca, Org. Lett., 2012, 14, 5014–5017; A. Porcheddu and L. De Luca, Adv. Synth. Catal., 2012, 354, 2949–2953; S. Gaspa, A. Porcheddu and L. De Luca, Org. Biomol. Chem., 2013, 11, 3803–3807.
16. In order to avoid possible decomposition of N-chloramine 2, it was recovered by removing the extraction solvent at room temperature and reduced pressure M. R. Monaco, P. Renzi, D. M. Scarpino Schietroma and M. Bella, Org. Lett., 2011, 13, 4546–4549 and references cited therein.
17. The previously optimised protocol for preparing imine 3 is not suitable for a one-pot procedure. Unfortunately, the formation of N-chloramine with NCS in polar protic solvents is accompanied by the formation of several by-products that significantly lower the yields and considerably complicate the final purification of N-chloramines.
18. Strict anhydrous conditions are generally required to prevent the hydrolysis of acid chlorides, ketenes and imines.
19. As expected, the ¹H-NMR spectrum of the crude reaction product showed the presence of a mixture of cis and trans isomers in a 92 : 8 ratio. These data fit those described in literature for the classic Staudinger reaction, where a cis selectivity is typically observed using acid chlorides activated with heteroatoms at the alpha position.
20. These results were not entirely unexpected, as the preparation of N-chloroanilines is extremely complex and N-chloroanilines are highly unstable compounds.
21. To our knowledge, there are relatively few literature precedents concerning the use of non-activated aliphatic acid chlorides in the synthesis of β-lactams M. Browne, D. A. Burnett, M. A. Caplen, L.-Y. Chen, J. W. Clader, M. Domalski, S. Dugar, P. Pushpavanam, R. Sher, W. Vaccaro, M. Viziano and H. Zhao, Tetrahedron Lett., 1995, 36, 2555–2558; C. M. L. Delpiccolo and E. G. Mata, Tetrahedron Lett., 2004, 45, 4085–4088.
22. Several unsuccessful attempts were conducted using NEt₃, ^tBuOK, DBA–DBU, pyridine and KO₂ as base.
23. SciFinder database.
24. G. I. Georg, Organic Chemistry of β-Lactams, Wiley VCH, New York, 1992; F. P. Cossio, A. Arrieta and M. A. Sierra, Acc. Chem. Res., 2008, 41, 925–936; Y. Liang, L. Jiao, S. Zhang, Z.-X. Yu and J. Xu, J. Am. Chem. Soc., 2009, 131, 1542–1549; H. Qi, X. Lia and J. Xu, Org. Biomol. Chem., 2011, 9, 2702–2714; J. Xu, ARKIVOC, 2009, 21–44; S.-L. Hou, X.-Y. Li and J. X. Xu, Sci. China: Chem., 2013, 56, 370–379; X. Z. Yang, N. Chen and J. Xu, J. Org. Chem., 2015, 80, 3611–3620.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra02744j

---