HETEROCYCLES

Communication | Regular issue | Vol. 83, No. 5, 2011, pp. 1013-1016
Received, 27th January, 2011, Accepted, 15th March, 2011, Published online, 18th March, 2011.
DOI: 10.3987/COM-11-12157

■ Unexpected Formation of Thiino[3’,4’:4,5]oxazolo[1,3-b][1,3]oxazepine Derivatives

Akikazu Kakehi,* Kennosuke Itoh, Hiroyuki Suga, Hirofumi Kobayashi, Hideyuki Muranaka, and Motoo Shiro

Department of Chemistry and Material Engineering, Faculty of Engineering, Shinshu University, 4-17-1 Wakasato, Nagano 380-8553, Japan

Abstract

The reactions of pyridinium 2-alkylthio-1-benzoyl-2-thioxoethylides with dimethyl acetylenedicarboxyate in THF at the reflux temperature afforded the unexpected products, dimethyl 2-alkylthio-4a-phenyl-4aH,5aH-thiino[3’,4’:4,5]oxazolo[1,3-b][1,3]oxazepine-3,4-dicarboxylates, together with dimethyl 10aH-4-alkylthio-5-benzoylpyrido[1,2-d][1,4]thiazepine-1,2- dicarboxylates.

We previously reported that the reactions of various pyridinium 2-alkylthio-2-thioxoethylides and dimethyl acetylenedicarboxyate (DMAD) afforded the corresponding dimethyl 4-alkylthio-10aH-pyrido- [1,2-d][1,4]thiazepine-1.2-dicarboxylates and/or their intramolecular Diels-Alder adducts2-4 and similar treatment of pyridinium (thiobenzoyl)aminides provided dimethyl 4-aryl-5-thia-2,3-diazatricyclo- [4.3.2.02,7]undeca-3,8,10-triene-6,11-dicarboxylates formed via the rearrangement of the initially generated dimethyl 2-aryl-5aH-pyrido[1,2-d][1,3,4]thiadiazepine-4,5-dicarboxylates.5,6 We were very interested in the latter rearrangement and tried the extension to several pyridinium 2-thioxoethylides in anticipation of similar rearrangement, but our effort was fruitless. We recently found the unexpected formation of some dimethyl 2-[2-acylthieno[2’,3’:2,3][1,4]thiazino[4,5-a]pyrrol-8-ylidene]succinates in the one-pot synthesis for 5-acylthieno[3,2-d]thiazole derivatives from the reactions of 5-acyl-3-(1-pyridinio)thiophene-2-thiolates with DMAD in refluxing xylene.7 This finding encouraged us to explore further this type of reaction, because the ring contraction from the pyridine ring to the pyrrole one in the structure of these products was distinctly confirmed. So, we reexamined the reaction of various pyridinium 2-thioxoethylides with DMAD by changing the reaction conditions and found the formation of the unexpected products, dimethyl 2-alkylthio-4a-phenyl-4aH,5aH-thiino[3’,4’:4,5]- oxazolo[1,3-b][1,3]oxazepine-3,4-dicarboxylates, in the reactions of pyridinium 2-alkylthio-1-benzoyl- 2-thioxoethylides with DMAD in nonstabilized tetrahydrofuran (THF).8
When the reactions of pyridinium 1-benzoyl-2-methylthio-2-thioxoethylides (1a, 1 mmol) with DMAD (2, 1.1 mmol) were heated in THF (20 mL) at the reflux temperature, a new type of product 3a, yellow prisms, mp 167–171 oC, ν (KBr) 1734 cm-1 (C=O), 1H-NMR (CDCl3): 2.40 (3H, s, SMe), 3.81 and 3.91 (each 3H, s, CO2Me), 5.11 (1H, br t, J = 8.2, 7.1 Hz, 8-H), 5.17 (1H, br t, J = 8.2, 8.1 Hz, 9-H), 6.02 (1H, s, 5a-H), 6.31 (1H, d, J = 7.1 Hz, 7-H), 7.30–7.35 (3H, m, Ph-H), 7.42 (1H, d, J = 8.1 Hz, 10-H), 7.57–7.63 (2H, m, Ph-H), 13C-NMR (CDCl3): 20.49, 53.04, 53.20, 88,66, 93.85, 104.26, 105.58, 110.24, 124.21, 126.88, 128.13, 129.15, 130.39, 133.48, 134.32, 136.23, 142.17, 162.19, 164.68, was obtained in 13% yield, together with dimethyl 5-benzoyl-4-methylthio-5aH-pyrido[1,2-d][1,4]thiazepine-1,2- dicarboxylates (4a, 13%). Similar treatment of pyridinium 2-methylthio- and 2-ethylthio-1- benzoyl-2-thioxoethylides (1b–d) with 2 provided 3b (6%)9 and 4b (9%), 3c (25%, mp 179–180 oC, ν (KBr) 1732 cm-1 (C=O), 1H-NMR (CDCl3): 1.82 (6H, s, 7-Me and 9-Me), 2.38 (3H, s, SMe), 3.81 and 3.90 (each 3H, s, CO2Me), 4.81 (1H, br s, 8-H), 6.06 (1H, s, 5a-H), 6.98 (1H, br s, 10-H), 7.28–7.35 (3H, m, Ph-H), 7.57–7.64 (2H, m, Ph-H), 13C-NMR (CDCl3): 20.75, 21.84, 21.90, 52.99, 53.13, 88.83, 91.74, 104.35, 109.85, 116.22, 119.15, 127.11, 127.99, 129.00, 130.75, 133.51, 135.10, 136.42, 150.21, 162.31, 164.78), and 3d (7%)9 and 4d (28%), respectively. These results are shown in Scheme 1.

The structures of 5aH-pyrido[1,2-d][1,4]thiazepines 4a–d were readily determined by spectral comparison with those of authentic samples prepared earlier by us.3 However, the structures of 3a–d were beyond our imagination for these reactions because the analyses of the chemical shifts and signal patterns in these 1H-NMR spectra precluded all of the structures which can be deduced from the reactions of 1a–d and 2. On the other hand, the elementary analyses and the high resolution mass spectra (HRMS)10 of 3a–d disclosed an interesting fact: that is, 3a–d did not have the composition of the 1 : 1 adduct between methylide 1 and DMAD 2 and the increase of another oxygen atom on them was suggested. Eventually, the structures, dimethyl 1-methylthio- or 1-ethylthio-4a-phenyl-4aH,5aH- thiino[3’,4’:4,5]oxazolo[1,3-b][1,3]oxazepine-3,4-dicarboxylates, for 3a–d were determined by the X-ray analysis (see Figure 1) of one compound 3a.11 As seen from this figure, products 3a–d had a 1,3-oxazepine skeleton to suggest the oxidative ring enlargement of the pyridine ring.

To explore the origin of the additional oxygen we examined the reactions of methylides 1a,c and 2 in stabilized and freshly purified THF at the reflux temperature and the reaction mixtures were analyzed by their 1H-NMR spectra. The reactions of 1a,c with 2 in THF stabilized by an antioxidant gave only pyridothiazepines 4a (20%) and 4c (36%). On the other hand, the reactions of 3a,c and 2 in THF purified by glass contour solvent systems provided 4a (37%) and 3c (7%) + 4c (40%) respectively. Interestingly, a small amount of oxazepine 3a was detected in the reactions of 1a with 2 in commercially available diethyl ether at the reflux temperature for 6h, but the similar reactions in CHCl3 and benzene did not afford any 3a at all.4 From these findings and the fact that THF or diethyl ether easily form the peroxide, the origin of the additional oxygen in products 3a–d can be considered to be peroxide. However, we have no idea for the formation mechanisms of 3a–d at present, since a reaction like this has no precedent, though the formation of 1,3-oxazepine derivatives in the photolyses of 2,6-disubstituted pyridine N-oxides is well known.12,13 Further investigation for the scope and the mechanism of this reaction is now in progress.

References

1. Preparation of new nitrogen-bridged heterocycles. 73. For part 72 of this series, see A. Kakehi, H. Suga, Y. Okumura, K Itoh, K. Kobayashi, Y. Aikawa, and K. Misawa, Chem. Pharm. Bull., 2010, 58, 1502.. CrossRef
2. A. Kakehi, S. Ito, and J. Hakui, Chem. Lett., 1992, 777. CrossRef
3. A. Kakehi, S. Ito, and J. Hakui, Bull. Chem. Soc. Jpn., 1993, 66, 3475. CrossRef
4. A. Kakehi, S. Ito, M. Mitani, and M. Kanaoka, Bull. Chem. Soc. Jpn., 1994, 67, 1646. CrossRef
5. A. Kakehi, S. Ito, F. Ishida, and Y. Tominaga, Heterocycles, 1995, 41, 2657. CrossRef
6. A. Kakehi, S. Ito, F. Ishida, and Y. Tominaga, J. Org. Chem., 1997, 62, 7788. CrossRef
7. A. Kakehi, H. Suga, Y. Okumura, and T. Nishi, Heterocycles, 2010, 81, 175. CrossRef
8. THF stored at room temperature without antioxidant for several years was employed in these reactions. 20 mL of the solvent left about 0.15g of residue after the evaporation of THF. The 1H-NMR spectrum of the residue showed complex signals and the presence and the quantity of THF peroxide could not be determined.
9. 3b: yellow prisms, mp 138–140 °C, ν (KBr) 1730 cm-1(C=O), 1H-NMR (CDCl3): 1.24 (3H, t, J = 7.2 Hz, SCH2CH3), 2.75 and 2.89 (each 1H, m, SCH2CH3), 3.81 and 3.91 (each 3H, s, CO2Me), 5.12 (1H, br t, J = 8.2, 7.1 Hz, 8-H), 5.16 (1H, br t, J = 8.2, 8.1 Hz, 9-H), 6.02 (1H, s, 5a-H), 6.31 (1H, d, J = 7.1 Hz, 7-H), 7.30–7.35 (3H, m, Ph-H), 7.55 (1H, d, J = 8.1 Hz, 10-H), 7.57–7.63 (2H, m, Ph-H). 3d: mp 152–154 °C, ν (KBr) 1734 cm-1 (C=O), 1H-NMR (CDCl3): 1.22 (3H, t, J = 7.3 Hz, SCH2CH3), 1.82 (3H, s, 7-Me or 9-Me), 1.83 (3H, s, 9-Me or 7-Me), 2.72 and 2.86 (each 1H, m, SCH2CH3), 3.81 and 3.90 (each 3H, s, CO2Me), 4.82 (1H, br s, 8-H), 6.06 (1H, s, 5a-H), 7.13 (1H, br s, 10-H), 7.28–7.35 (3H, m, Ph-H), 7.57–7.64 (2H, m, Ph-H).
10. The compounds 3a–d gave satisfactory elementary analyses and the HRMS data are as follows: 3a, C21H20NO6S2 (M+H)+: Calcd; 446.0727 (Found; 446.0729). 3b, C22H22NO6S2 (M+H)+: Calcd; 460.0883 (Found; 460.0893). 3c, C23H24NO6S2 (M+H)+: Calcd; 474.1040 (Found; 474.1039). 3d, C24H26NO6S2 (M+H)+: Calcd; 488.1196 (Found; 488.1190).
11. A yellow crystal (0.20 x 0.14 x 0.10 mm) grown from CHCl3-hexsane was mounted on a glass fiber. All measurements were made on a Rigaku RAXIS RAPID imaging plate area detector with graphite-monochromated CuKα radiation (λ = 1.54187 Å). Crystal data of this compound are as follows: 3a: C21H20N2O2S2; M = 445.50; triclinic, space group P-1 (#2), Z = 2 with a = 8.71314 (16) Å, b = 9.23899 (17) Å, c = 14.5356 (10) Å, α = 87.223 (6)°, β = 85.705 (6)°, γ = 62.075 (4)°; V = 1030.86 (8) Å3, and Dcalc.= 1.435 g/cm3. The structure was solved by a direct method (SHELX97).18 The non-hydrogen atoms were refined anisotropically. Hydrogen atoms were refined using the riding model. The final R1- and wR2-factors after full-matrix least-squares refinements were 0.0325 for (I>2.00σ(Ι)) and 0.0857 for all observed reflections (3699).
12. M. Ishikawa, C. Kaneko, I. Yokoe, and S. Yamada, Tetrahedron, 1969, 25, 295. CrossRef
13. O. Buchardt, C. L. Christian, and N. Harrit, J. Org. Chem., 1972, 37, 3592. CrossRef