Catalytic conversion of lactide to optically pure heterocycles

Abstract

The catalytic preparation of optically pure nitrogen heterocycles was developed via cycloaddition using lactide catalyzed by tin alkoxide.

---

Glycolide and lactide, the related cyclic dimers of glycolic and lactic acid, are plant-derived substrates. Of special interest is the fact that lactide is an easily available chiral source. The major utility of these substrates is in the application to biodegradable aliphatic polyesters such as poly(glycolide) (PGA) and poly(lactide) (PLA).¹ As far as we could ascertain, however, their direct transformation to monomeric fine chemicals such as heterocycles has not been reported.² Oxazolidine-2,4-diones are used in the areas of medicinal,³ agricultural⁴ and synthetic applications.⁵ We have recently reported the equimolar tin-mediated synthesis of heterocycles (Scheme 1).⁶ The in situ formed tin alkoxide was added to an isocyanate followed by cyclization of the remaining functional group. Thus α-stannoxy-ester A and stannyl amine B were the reactive intermediates. We present here the catalytic synthesis of oxazolidine-2,4-diones from glycolide and lactide by using a tin alkoxide catalyst. In the catalytic reaction, A and B worked well as active catalytic species.^7,8 In particular, optically pure oxazolidine-2,4-diones 4 were obtained directly from easily available chiral sources.

Scheme 1 The equimolar tin-mediated reaction.

Initially, methyl glycolate (1a) was used as the substrate as we thought that it could be a good precursor of α-stannoxy-ester A. Table 1 shows the results of the catalytic effect in the reaction with BuN=C=O.⁹ Without a catalyst, the reaction proceeded in a low yield even with heating where no cyclization proceeded, and linear adduct 2a′ was obtained in a 10% yield (entry 1). In the presence of a tin catalyst (10 mol%) such as di-n-butyltin oxide (Bu₂Sn=O) or the tin(II) bis(2-ethylhexanoate) system [Sn(Oct)₂–MeOH],¹⁰1a reacted with BuN=C=O well at rt for 3 h, however, only linear adduct 2a′ was obtained (entries 2 and 3).¹¹ On the other hand, using tri-n-butyltin methoxide (Bu₃SnOMe) as the catalyst afforded a cyclic product, 1,3-oxazolidin-2,4-dione 2a (entry 4). Further, di-n-butyltin dimethoxide [Bu₂Sn(OMe)₂] was employed as an effective catalyst for the preparation of 2a (entry 5). Microwave irradiation gave the desired 2a in only 10 min (entry 6).

Table 1 Reaction of methyl glycolate with BuN=C=O^a

Entry	Catalyst	Conditions^b	2a [%]^c	2a′ [%]
a 1a 1 mmol, BuN=C=O 1 mmol, cat. 0.1 mmol, MeCN 1 mL. b rt = room temperature; MW = microwave. c tr = trace.
1	None	80 °C, 3 h	tr	10
2	Bu2Sn=O	rt, 3 h	tr	72
3	Sn(Oct)2–2 MeOH	rt, 3 h	tr	99
4	Bu3SnOMe	rt, 3 h	19	60
5	Bu2Sn(OMe)2	rt, 3 h	70	6
6	Bu2Sn(OMe)2	MW, 90 °C (30 W), 10 min	80	4

---

A plausible catalytic cycle is shown in Scheme 2. The Sn–O and Sn–N bonds bear high nucleophilicity.¹⁰ Initially, tin methoxide reacts with 1a to give α-stannoxy-ester A. The Sn–O bond of A reacts with isocyanate to form stannyl-carbamate B.¹² The resultant Sn–N bond of B reacts with the remaining ester moiety. The Sn–O bond of cyclized product C reacts with the starting 1a. As a result, 1,3-oxazolidine-2,4-dione 2a is obtained and the catalytic active species A is regenerated.¹³ On the other hand, linear adduct 2a′ is obtained by the protonolysis of B with 1a. Bu₂Sn(OMe)₂ forms a dimeric structure in which apical Sn–OMe bonds of 5-coordinated tin have high nucleophilicity.¹⁴ Thus, the highly nucleophilic Bu₂Sn(OMe)₂ is employed as an effective catalyst rather than the Sn(Oct)₂–MeOH-system.¹⁵

Scheme 2 A plausible catalytic cycle in the reaction using 1a.

Thus, protonation occurs instead of cyclization. In fact, when glycolic acid was used as a starting material, no desired products were obtained when using Bu₂Sn(OMe)₂ as the catalyst.

As shown in Table 2, we then tried to use the related cyclic dimer, glycolide 1b, as a substrate. This substrate has no proton source. When the reaction with BuN=C=O was carried out under the optimized conditions in Table 1, entry 6, the reaction scarcely proceeded (entry 1). However, the reaction using higher MW power increased the yield of 2a (entries 2 and 3), and the reaction at 120 °C for 20 min gave a quantitative yield (entry 4). The reaction with a smaller amount of catalyst decreased the yield (entry 5). The polyester, PGA, can be obtained by the ring-opening polymerization of glycolide 1b.¹ The most widely used catalyst for industrial preparation is the tin(II) bis(2-ethylhexanoate) system [Sn(Oct)₂–MeOH]. This catalytic system did not give 2a from 1a as shown in Table 1, entry 3, however, using glycolide 1b as a substrate enabled the preparation of 2a (entry 6). Thus, the absence of proton sources in the reaction of 1b allowed the use of Sn(Oct)₂ as the catalyst. The reaction using 1b was highly atom-economical and no side products were obtained. In place of aliphatic isocyanate, aromatic isocyanates were reactive giving oxazolidine-2,4-diones, 2b and 2c (entries 7–10).

Table 2 Reaction of glycolide 1b with isocyanates^a

Entry	R	Catalyst	Conditions	Product 2 [%]
a 1b 0.5 mmol, isocyanate 1 mmol, tin catalyst 0.1 mmol, MeCN 1 mL. b Sn(Oct)2=Sn(OCOC7H17)2.
1	Bu	Bu2Sn(OMe)2	MW 80 °C (30 W), 5 min	2a	trace
2		Bu2Sn(OMe)2	MW 95 °C (50 W), 10 min		11
3		Bu2Sn(OMe)2	MW 110 °C (100 W), 10 min		78
4		Bu2Sn(OMe)2	MW 120 °C (100 W), 20 min		99
5		Bu2Sn(OMe)2 (0.1 equiv)	MW 120 °C (100 W), 20 min		44
6		Sn(Oct)2–2 MeOH^b	MW 120 °C (100 W), 20 min		74
7	PMP	Bu2Sn(OMe)2	MW 120 °C (100 W), 20 min	2b	99
8		Sn(Oct)2–2 MeOH	MW 120 °C (100 W), 20 min		71
9	p-IC6H4	Bu2Sn(OMe)2	MW 120 °C (100 W), 20 min	2c	79
10		Sn(Oct)2–2 MeOH	MW 120 °C (100 W), 20 min		86

---

Hence, organotin alkoxides could be good candidates to catalyze the cycloaddition of 1b in which the high nucleophilicity of the Sn–O bond to the lactone ring is an important factor.¹⁶ A plausible catalytic cycle is shown in Scheme 3. Initially, tin methoxide cleaves the acyl–oxygen bond of 1b. The resultant Sn–O bond of A′ reacts with an isocyanate to form stannyl-carbamate B′. The resultant Sn–N bond reacts with the remaining carbonyl moiety intramolecularly to give 1,3-oxazolidine-2,4-dione 2. The subsequent addition–cyclization through A and B also gave product 2, regenerating tin methoxide.

Scheme 3 A plausible catalytic cycle in the reaction using 1b.

As shown in Table 3, instead of isocyanates, when isothiocyanate (BuN=C=S) was used, 1,3-oxazolidine-2-thione 3a, was obtained catalyzed by Bu₂Sn(OMe)₂ (entry 1). On the other hand, the Sn(Oct)₂–2MeOH catalytic system gave no satisfactory results (entry 2). Using other isothiocyanates also gave the corresponding adducts 3b, 3c, and 3d that were effectively catalyzed by Bu₂Sn(OMe)₂ (entries 3–5), whereas, the Sn(Oct)₂–2MeOH system basically showed less catalytic activity (entries 3–5, parentheses). It is known that a Sn–OR bond adds across the C=S group of an isothiocyanate (R′N=C=S) to give a Sn–S–C(=NR′)OR adduct.¹⁷ Because of the high affinity of Sn toward the S atom,¹⁸ the cyclization step accompanying the cleavage of the Sn–S bond would be difficult compared with the reaction using isocyanates (Table 2).

Table 3 Reaction of glycolide 1b with isothiocyanates^a

Entry	R	Catalyst	Product 3 [%]
a 1b 0.5 mmol, isothiocyanate 1 mmol, tin catalyst 0.1 mmol, MeCN 1 mL. b Yields from the Sn(Oct)2–2MeOH system-catalyzed reactions are in parentheses.
1	Bu	Bu2Sn(OMe)2	3a	91
2		Sn(Oct)2–2MeOH		12
3	Ph	Bu2Sn(OMe)2	3b	> 99 (tr)^b
4	PMP	Bu2Sn(OMe)2	3c	90 (tr)^b
5	p-NO2C6H4	Bu2Sn(OMe)2	3d	90 (25)^b

---

Lactide 1c is also a naturally occurring compound like glycolide 1b, and is an easily available chiral source.¹ As shown in Scheme 4, the reaction of 1c with p-methoxyphenyl isocyanate proceeded quantitatively catalyzed by Bu₂Sn(OMe)₂ at 130 °C for 10 min (Scheme 4, (1)). However, after product 4a was isolated, its enantiopurity decreased to 56% ee despite the fact that optically pure 1c was used. Namely, racemization occurred during the reaction. It has already been reported that tin amide (Bu₃Sn–NR₂) provides tin enolate from reaction with α-alkoxyketone.¹⁹ Hence, as shown in Scheme 5, intermediate stannyl-carbamate C had enough basicity to abstract acidic hydrogen to cause undesirable racemization. Fortunately, when the Sn(Oct)₂–MeOH system was used as the catalyst, 4a was obtained with no loss of enantiopurity (Scheme 4, (2)). As described, the basicity of the Sn–N bond generated in the Sn(Oct)₂–MeOH system was lower than that of Bu₂Sn(OMe)₂. So abstraction of the acidic hydrogens and the inducement of racemization was suppressed.

Scheme 4 Reaction of lactide with isocyanate.

Scheme 5 A plausible racemization mechanism.

Table 4 shows the results of the reactions of lactide 1c with various isocyanates catalyzed by the Sn(Oct)₂–MeOH system. Thus, optically pure products, 4a–4c, were obtained (entries 1–3). The reaction under MW irradiation was essential because reflux conditions did not give the desired product 4a (entry 1, parenthesis). Phenyl lactide 1d is also easily available in an optically pure form.²⁰ Thus 5-benzyl-1,3-ozazolidine-2,4-diones 4d–4f were obtained with no loss of enantiopurity (entries 4–6).

Table 4 Reaction of lactide derivatives 1b or 1c with isocyanates^a

Entry	R^1		R^2^b	Conditions	Product 4 [%]		ee [%]
a 1c or 1d 0.5 mmol, isocyanate 1 mmol, Sn(Oct)2 0.1 mmol, MeOH 0.2 mmol, MeCN 1 mL. b PMP = p-MeOC6H4; Ts = tosyl. c 80 °C, 3 h.
1	Me	1c	PMP	MW 155 °C (150 W), 30 min	4a	99	99 (trace)^c
2	Me	1c	p-IC6H4	MW 155 °C (150 W), 30 min	4b	96	98
3	Me	1c	Ts	MW 160 °C (150 W), 30 min	4c	89	98
4	Bn	1d	PMP	MW 155 °C (150 W), 30 min	4d	99	94
5	Bn	1d	p-IC6H4	MW 158 °C (150 W), 30 min	4e	90	97
6	Bn	1d	Ts	MW 160 °C (150 W), 30 min	4f	79	92

---

Conclusions

In conclusion, the first synthetic use of plant-derived substrates, glycolide and lactide, to give heterocycles was achieved by using tin alkoxide catalysts. In particular, optically pure products were obtained from easily available chiral sources.

References

1. For reviews see: (a) O. D. Cabaret, B. M.-Vaca and D. Bourissou, Chem. Rev., 2004, 104, 6147–6176; (b) R. H. Platel, L. M. Hodgson and C. K. Williams, Polym. Rev., 2008, 48, 11–63.
2. Alcoholysis: (a) A. Grala, J. Ejfler, L. B. Jerzykiewicz and P. Sobota, Dalton Trans., 2011, 40, 4042–4044; (b) K. Phomphrai, S. Pracha, P. Phonjanthuek and M. Pohmakotr, Dalton Trans., 2008, 3048–3050; Aminolysis: (c) G. A. Brine, K. G. Boldt, D. Prakash, D. J. Kotchmar, V. C. Bondeson, D. J. Bradley, P. Singh and F. I. Carroll, J. Chem. Soc., Perkin Trans. 1, 1991, 1809–1814; (d) Bromination: F. Jing and M. A. Hillmyer, J. Am. Chem. Soc., 2008, 130, 13826–13827; (e) Alkylation: K. C. Nicolaou, R. Jautelat, G. Vassilikogiannakis, P. S. Baran and K. B. Simonsen, Chem.–Eur. J., 1999, 5, 3651–3665; (f) Reduction: E. Balaraman, E. Fogler and D. Milstein, Chem. Commun., 2012, 48, 1111–1113; (g) R. J. Wehmschulte and L. Wojtas, Inorg. Chem., 2011, 50, 11300–11302.
3. Hypoglycemic agents: (a) R. L. Dow, B. M. Bechle, T. T. Chou, D. A. Clark, B. Hulin and R. W. Stevenson, J. Med. Chem., 1991, 34, 1538–1544; (b) NK₁ antagonists: R. T. Lewis, A. M. Macleod, K. J. Merchant, F. Kelleher, I. Sanderson, R. H. Herbert, M. A. Cascieri, S. Sadowski, R. G. Ball and K. Hoogsteen, J. Med. Chem., 1995, 38, 923–933.
4. Fungicide: (a) J. A Sternberg, D. Geffken, J. B Adams Jr, R. Pöstages, C. G Sternberg, C. L Campbell and W. K Moberg, Pest Manage. Sci., 2001, 57, 143–152; (b) Y.-J. Zheng, R. Shapiro, W. J Marshall and D. B Jordan, Bioorg. Med. Chem. Lett., 2000, 10, 1059–1062; (c) D. B Jordan, R. S Livingston, J. J. Bisaha, K. E. Duncan, S. O. Pember, M. A. Picollelli, R. S. Schwartz, J. A Sternberg and X.-S. Tang, Pestic. Sci., 1999, 55, 105–118.
5. (a) J. W. Clark-Lewis, Chem. Rev., 1958, 58, 63–99; (b) G. Galliani, B. Rindone and F. Saliu, Tetrahedron Lett., 2009, 50, 5123–5125.
6. I. Shibata, R. Kojima, S. Tsunoi, T. Nozaki, T. Watanabe, A. Ninomiya, M. Yasuda and A. Baba, Org. Biomol. Chem., 2010, 8, 2009–2011.
7. Synthetic uses of organotin compounds, see: (a) A. Baba, I. Shibata and M. Yasuda, in Comprehensive Organometallic Chemistry III, ed. P. Knochel, Elsevier, Oxford, 2007, vol. 9, ch. 8, pp. 341–380.; (b) M. Pereyre, P. J. Quintard and A. Rahm, in Tin in Organic Synthesis, Butterworth, London, 1987, pp. 261–258; (c) A. G. Davies, in Organotin Chemistry, VCH, Weinheim, 1997, pp. 166–193..
8. Sn–O and Sn–N bonds bear high nucleophilicity. In some cases, their nucleophilicity is higher than that of the corresponding free alcohols and amines: E. W. Abel, D. A. Armitage and D. B. Brady, Trans. Faraday Soc., 1966, 62, 3459–3462.
9. We have recently reported the reaction of α-hydroxy ketones with isocyanate catalyzed by Bu₂Sn(OMe)₂: R. Kojima, S. Sawamoto, A. Okamura, H. Takahashi, S. Tsunoi and I. Shibata, Eur. J. Org. Chem., 2011, 7255–7258.
10. (a) H. R. Kricheldorf, I. Kreiser-Saunders and A. Stricker, Macromolecules, 2000, 33, 702–709; (b) A. Kowalski, A. Duda and S. Penczek, Macromolecules, 2000, 33, 7359–7370.
11. Tin methoxide catalyzed urethane formation: J. Otera, T. Yano and R. Okawara, Chem. Lett., 1985, 14, 901–904.
12. A. G. Davies, T. N. Mitchell and W. R. Symes, J. Chem. Soc. (C), 1966, 1311–1315.
13. Tin-catalyzed reaction using epoxides: I. Shibata, A. Baba, H. Iwasaki and H. Matsuda, J. Org. Chem., 1986, 51, 2177–2184.
14. A. G. Davies and P. G. Harrison, J. Chem. Soc. (C), 1967, 298–300.
15. In the Sn(Oct)₂–MeOH-system, the catalytic species form 4-coordinated tin,^1a and has no alkyl substituents on tin. Hence compared with the Bu₂Sn(OMe)₂ system, the Sn–N bonds in B would not bear enough nucleophilicity toward an ester moiety beyond protonolysis.
16. Cleavage of lactide by tin alkoxide: M. H. Chisholm and E. E. Delbridge, New J. Chem., 2003, 27, 1167–1176.
17. I. Shibata, A. Baba, H. Kashiwagi and H. Matsuda, Bull. Chem. Soc. Jpn., 1986, 59, 341–343.
18. (a) K. Itoh, I. K. Lee, I. Matsuda, S. Sakai and Y. Ishii, Tetrahedron Lett., 1967, 8, 2667–2670; (b) K. Itoh, Y. Fukumoto and Y. Ishii, Tetrahedron Lett., 1968, 9, 3199–3202.
19. I. Shibata, K. Yasuda, Y. Tanaka, M. Yasuda and A. Baba, J. Org. Chem., 1998, 63, 1334–1336.
20. T. L. Simmons and G. L. Baker, Biomacromolecules, 2001, 2, 658–663.

---

Footnote

† Electronic Supplementary Information (ESI) available. See DOI: 10.1039/c2ra20831h/

---