Open Access

[3 + 2] Cycloaddition reactions of thioisatin with thiazolidine-2-carboxylic acid: a versatile route to new heterocyclic scaffolds

  • Sonali Verma1,
  • Johnson George1,
  • Saurabh Singh1,
  • Pushpa Pardasani1 and
  • Ramchand Pardasani1Email author
Organic and Medicinal Chemistry Letters20111:6

DOI: 10.1186/2191-2858-1-6

Received: 18 March 2011

Accepted: 6 September 2011

Published: 6 September 2011

Abstract

A facile synthesis of azabicycloadducts is described by 1,3-dipolar cycloaddition reactions of thioisatin with thiazolidine-2-carboxylic acid in the presence of various electron rich and electron deficient dipolarophiles. Theoritical calculations have been performed to study the regioselectivity of products. The geometrical and energetic properties have been analyzed for the different reactants, transition states and cycloadducts formed.

Keywords

Azabicycloadducts 1,3-Dipolar cycloaddition reactions AM1 Calculations Thioisatin Thiazolidine-2-carboxylic Acid

Background

The construction of sophisticated molecules requires viable, selective and highly reliable reactions as potent synthetic tools [13]. The 1,3-dipolar cycloaddition [49] has also become one of the most important legation method in biology and material chemistry. Thioisatin derivatives [10] have received the attention of biochemists because of their therapeutic and biological activities. Similarly thiazolidine-2-carboxylic acid [11, 12] exhibit strong antioxidant properties. Therefore any heterocyclic scaffold containing these two moieties might be expected to have considerably enhanced biological activities.

The 1,3-dipolar cycloaddition reaction of an azomethine ylide with an alkene leads to the formation of pyrrolidine [13, 14] derivatives. Recently, we have reported the results on azomethine ylides derived from 9,10-phenanthrenequinone and some secondary cyclic α-amino acids with different dipolarophiles [15]. Herein we report the reactivity and regioselectivity of 1,3-dipolar cycloaddition reactions of azomethine ylides derived from benzo[b]thiophene-2,3-dione with thiazolidine-2-carboxylic acid in the presence of various acetylenic and ethylenic dipolarophiles. Besides synthetic work, a systematic and comprehensive theoretical study at Gaussian 03 [16] suite of programs has been carried out to address the mechanism as well as regio- and stereochemical course of the reactions.

2. Result and discussion

The reaction of thioisatin 1 with thiazolidine-2-carboxylic 2 acid was carried out in equimolar ratio in refluxing dry acetonitrile in the presence of diphenylacetylene as dipolarophile to afford a diastereomeric mixture of {(5R,8R)-spiro-6,7-diphenyl-1aza-4thia-bicyclo [3,3,0]-6-octene-8,3'}-benzo[b]thiophene-2'-one} 8 as cis/trans isomers. Analogous reactions of thioisatin with other dipolarophiles viz methyl acrylate, phenylacetylene, phenylpropyne and ethyl phenyl propiolate produced diasteroisomeric mixtures of cycloadducts 5-9 in 75%- 63% yield. The mechanism for the formation of the cycloadducts 5-9 involve the initial formation of an iminium species 3 followed by the loss of CO2 via stereospecific 1,3-cycloreversion [17] to azomethine ylides 4. Subsequent [3 + 2] cycloaddition with various dipolarophiles then produce novel azabicycloadducts (Scheme 1).
Scheme 1

Reaction of thioisatin with thiazolidine-2-carboxylic acid in presence of different dipolarophiles produced regioisomeric mixtures of cycloadducts.

The structure of all the cycloadducts has been ascertained from their spectral data. Thus the IR spectrum of a typical diphenylacetylene cycloadducts 8 showed characteristic bands at 1715, 1420 and 690 for > C = O, C-N and C-S stretching vibrations respectively. Its 1H NMR spectrum showed a triplet at δ 2.51 (J = 2.7Hz) for 3H protons, another triplet at δ 2.60 (J = 3.0Hz) was associated with 2H protons, a singlet at δ 4.15 appeared for 5H and the aromatic protons resonated between δ 6.79- δ 8.12 ppm. Its 13C NMR spectrum showed a signal at 183.54 for C-2' carbonyl carbon, aromatic carbons appeared in the range δ 146.41-δ131.32 ppm, the olefenic carbons (C-6, C-7) resonated at δ 127.32 and 126.54, spiro carbon (C-8) was noticed at δ 86.23, C-5 at δ 46.72, C-2 at δ 45.43, and C-3 at 34.56 ppm respectively. Additional evidence was gathered from the mass spectrum of cycloadduct 8. The molecular ion and the base peaks were present at m/z 413(32%) and 235(100%) respectively; another peak at m/z 385(39%) corresponded to [M-CO].+ whereas the peak at m/z 108(35%) was assigned to [C6H4S].+.

2.1 Theoretical calculations: Regioselectivity of cycloadducts 5-9

The stereochemical course of the cycloaddition was examined by AM1 calculations. To calculate the relevant activation and stabilization energies, minimized geometries of the reactants, products and transition states are required. The molecular geometry of the simplest azomethine ylide (abbreviated as am y) 4 derived from thioisatin and thiazolidine-2-carboxylic acid has been optimized on Gaussian 03 program at AM1 level.

Geometry optimization showed that amy 4 has almost planar structure (Figure 1). Instead of having an envelope shape the thiazolidine ring is planar and lies in the same plane as that of thioisatin ring. It may exist in two isomeric forms, one in which > C = O group and C-H of the dipole are syn to each other, 4 syn , and the other in which these two groups are anti, 4 anti (Figure 2).
Figure 1

Optimized geometry of amy 4 anti.

Figure 2

Mode of attack of dipolarophile (diphenylacetylene) on amy 4.

Methyl acrylate may approach either of the amy with the formation of products having three chiral centers. Therefore a total 8 + 8 = 16 isomers 5a-5p are possible (Figure 3).
Figure 3

Possible isomers of cycloadduct 5 and its regioselectivity.

Attack of methyl acrylate on syn amy results in the inward movement of thiazolidine ring towards thioisatin nucleus which imposes steric hindrance and makes it unstable. In fact we failed to locate the transition state in any such case 5i-5p. The remaining 8 isomers may be obtained by the attack of methyl acrylate on anti amy. Out of these 8 possibilities only four 5a-5d have concerted mechanism. We have optimized the geometries of all the four isomers. Results show that all isomers have almost same ΔH f, indicating that thermodynamically all are nearly equally stable.

We have carried out transition state calculations on all the four isomers but have been successful in locating the transition state for only two isomers 5a-5b (Figure 4).
Figure 4

Transition states of cycloadducts 5a and 5b.

The transition state of the concerted 1,3-dipolar cycloadditions is usually controlled by frontier molecular orbitals of dipolarophiles and dipole (azomethine ylide). The favoured path involves HOMOdipole and LUMOdipolarophile. The ΔH f , HOMO, LUMO energies and HOMO-LUMO energy gaps of azomethine ylides 4 with dipolarphiles are given in Table 1.
Table 1

ΔHf, HOMO, LUMO, energies and H-L And L-H energy gaps

Cycloadduct

5a

5b

5c

5d

ΔH f (Kcal/mol)

81.34

82.98

82.86

82.68

From the Table, it may be concluded that HOMOdipole-LUMOdipolarophile energy gap is lower than the LUMOdipole-HOMOdipolarophile gap and therefore the dominant FMO approach is HOMOdipole-LUMOdipolarophile. Both the HOMO and the LUMO of the dipole show uneven distribution of the electron density along the C-N-C dipole. In the HOMO case, the orbital coefficient is larger at C1 (0.288) than at C2 (-0.163). Similarly in the LUMO of methyl acrylate the atomic orbital coefficient are (0.611) and (-0.521) respectively (Figure 3). Thus it may be concluded that there is better overlap when -COOMe group lies towards thioisatin ring, giving two possibilities 5a and 5b. Out of these 5a is formed in diastereomeric excess probably due to the endo approach of -COOMe.

2.2 Regioselectivity of the addition of symmetrical and unsymmetrical acetylenes

Parallel to methyl acrylate, diphenylacetylene can also attack either of the azomethine ylide (4 syn or 4 anti ) with the formation of products having two chiral centers. Therefore a total of 4 + 4 = 8 isomers (4 pairs of enantiomers) could be possible 8a-8h (Figure 5).
Figure 5

Possible isomers of cycloadduct 8.

Attack of diphenylacetylene on syn-azomethine ylide 4 syn results in the inward movement of the thiazolidine ring towards the thioisatin nucleus and the transition state could not be located even in a single case (Figure 2) ruling out the possibility of the formation of products 8e-8h. Thus it leaves the possibility of attack on the anti azomethine ylide 4 anti and hence only four isomers 8a-8d are left for consideration. We have optimized the geometry of all the four isomers. Results show that all isomers have almost same ΔH f, indicating that thermodynamically all are nearly equally stable.

Of remaining four possibilities 8c and 8d where N and H atoms on the adjacent carbon atoms do not lie on the same side, the transition state could not be located because concerted mechanism is not possible in such a situation. This leaves only two isomers 8a-8b for consideration. Out of these two isomers we could optimize the transition state in case of 8a only (Figure 6). This can be explained using the FMO approach along with the endo approach of the phenyl ring (Figure 7) as discussed above.
Figure 6

Transition state of diphenylacetylene cycloadduct 8.

Figure 7

Atomic orbital coefficients and overlapping of dipole( amy ) with diphenylacetylene.

Similarly attack of unsymmetrical dipolarophile, such as ethyl phenyl propiolate, on syn or anti azomethine ylide may produce a cycloadduct having two chiral centres and therefore a total of 4 + 4 = 8 stereoisomers 9a-9h could be possible (Figure 8) and it was concluded that isomer 9a is formed preferentially (Figure 9). The energy profile diagrams for azabicycloadducts 5-9 are presented in (Figure 10).
Figure 8

Possible isomers of cycloadduct 9.

Figure 9

Atomic orbital coefficients and overlapping of amy with ethylphenyl propiolate.

Figure 10

Energy profile diagrams of azabicycloadducts 5-9 (all values in Kcal/mol).

ΔH f -R, ΔH f -Ts, ΔH f -P, Ea(activation energy) and stabilization energy of amy with different dipolarophiles have been tabulated in Table 2.
Table 2

ΔHf-R, ΔHf-Ts, ΔHf-P, Ea. and stabilization energy of amy with different dipolarophiles

Cycloadduct

8a

8b

8c

8d

ΔHf (Kcal/mol)

109.38

112.32

109.86

110.38

3. Conclusions

From the above discussions, it may be concluded that:
  1. a.

    The azomethine ylide exits in two conformations; 4 syn and 4 anti .

     
  2. b.

    The dominant FMO approach is HOMOdipole-LUMOdipolarophile as this energy gap is lower than the LUMOdipole-HOMOdipolarophile gap.

     
  3. c.

    Azomethine ylide is stabilized by the delocalization of dipolar charge into thioisatin nucleus, thus increasing the activation energy for the reaction path.

     

4. Experimental

The uncorrected melting points were taken in open glass capillaries. The IR spectra were recorded on a Nicolet Magna IR Spectrometer Model 550 in KBr pellets and band positions are reported in wave numbers (cm-1). The 1H NMR spectra and 13C NMR spectra have been recorded on a Bruker DRX-300 MHz and 75.47 MHz model respectively in CDCl3 and DMSO using tetramethylsilane as an internal standard. The chemical shifts are given in δ ppm values. The mass spectra were recorded on a JEOL-SX 102 (FAB). Most of the spectra were recorded at CDRI, Lucknow, India. Elemental analyses were performed on a Perkin Elmer Series C, H, N, S Analyzer 2400. The solvents were purified by standard procedures [18, 19]. Acetonitrile was dried by refluxing with anhydrous calcium chloride for 5-6 h and then distilling it. Column chromatography was performed on silica gel 60 (Merck).

Methods

Synthesis of (5S,7R,8R)-spiro-{7-methoxycarbonyl-1-aza-4-thia-bicyclo [3, 3, 0]-octane 8, 3'}-benzo[b]thiophene-2'-one (5)

An equimolar mixture of thiosatin 1 (0.36 gm, 2.0 mmol), thiazolidine-2-carboxylic acid 2 (0.26 gm, 2.0 mmol) and methyl acrylate (0.32 gm, 2.0 mmol) in dry acetonitrile (50 ml) was refluxed for 22 h. The reaction was monitored by TLC until the consumption of the reactants. The reaction mixture was filtered, solvent evaporated and the residue was subjected to column chromatography. The petroleum ether/chloroform (4:1) fraction afforded the desired azabicycloadduct 5 as pale brown solid.

Pale brown solid, yield: 0.33g (70%), mp: 105-107°C. IR (KBr): 1710(C = O),1450(C-N), 715(C-S) cm-1. 1H NMR (CDCl3, δ ppm): = 2.29 (1H, t, J = 3.0Hz 7-CH), 2.46(2H, dd, J 1 = 4.2Hz, J 2 = 3.3Hz 6-CH2), 2.50(2H, t, J = 1.2Hz 3-CH2), 2.68(2H, t, J = 2.1Hz 2-CH2), 3.10(3H, s, OCH3), 4.15(1H, t, J = 3.54 Hz, 5-CH), 7.36-7.54(4H, m, ArH). 13C NMR (CDCl3, δ ppm): 35.43(CH2S), 44.58(CH2), 46.34(CH), 85.99(C-N), 125.86-124.32(C = C), 142.58-131.36(CHaro), 175.67(O = C-O), 184.32(C = O). EI-MS: m/z (%) = 321(M+ 38), 275(100), 261(20), 284(15), 234(18). Anal. Calcd. For C15H15NO3S2: C, 56.05%; H, 4.70%; N, 4.36%. Found: C, 56.45%; H, 4.78%; N, 4.76%.

(5R,8R)-spiro-{7-phenyl-1-aza-4-thia-bicyclo [3,3,0]-6-octene-8,3'}-benzo[b] thiophene-2'-one (6)

Coffee brown powder, yield: 0.31g (65%), mp: 95-97°C. IR (KBr): 1720(C = O), 1445(C-N), 690(C-S) cm-1. 1H NMR (CDCl3, δ ppm): = 2.50(2H, t, J = 3.0 Hz, 3-CH2), 2.68(2H, t, J = 3.3Hz, 2-CH2), 4.12(1H, d, J = 2.4Hz, 5-CH), 4.32(1H, d, J = 6.6Hz, 6-CH), 7.32-7.54(9H, m, ArH). 13C NMR (CDCl3, δ ppm): 31.51(CH2S), 36.54(CH2), 45.46(CH), 85.95(C-N), 128.81-124.49(C = C), 143.87-132.36(CHaro), 180.79(C = O). EI-MS: m/z (%) = 337(M+ 36), 203(100), 309(20), 108(34). Anal.Calcd for C19H15NOS2: C, 67.62%; H, 4.48%; N, 4.15%. Found: C, 67.78%; H, 4.53%; N, 4.32%.

(5R,8R)-spiro-{7-phenyl-6-methyl-1-aza-4-thia-bicyclo [3,3,0]-6-octene-8,3'}- benzo[b]thiophene-2'-one (7)

Brownish solid, yield: 0.32g (68%), mp: 120-122°C. IR (KBr): 1725(C = O), 1420(C-N), 678(C-S) cm-1.1H NMR (CDCl3, δ ppm): 2.14(3H, d, J = 7.2 Hz, Me), 2.50(2H, t,J = 3.0Hz 3-CH2), 2.68(2H, t,J = 3.2Hz 2-CH2), 2.84(1H, q, J = 3.6Hz,5-CH), 7.59-8.45(9H, m, ArH). 13C NMR(CDCl3, δ ppm): 32.33(CH2S), 35.43(CH2), 44.45(CH), 86.54(C-N), 124.23-122.86(C = C), 143.41-130.63(CHaro), 184.34(C = O). Anal.Calcd for C20H17NOS2: C, 68.34%; H, 4.87%; N, 3.98%. Found: C, 68.55%; H, 4.93%; N, 4.13%.

(5R,8R)-spiro-{6,7-diphenyl-1-aza-4-thia-bicyclo [3,3,0]-6-octene 8, 3'}- benzo[b]thiophene-2'-one (8)

Shiny brown powder, yield: 0.29g (63%), mp: 135-133°C. IR (KBr): 1715(C = O), 1420(C-N), 6905(C-S) cm-1. 1H NMR (CDCl3, δ ppm): 2.51(2H, t,J = 2.7Hz, 3-CH2), 2.60(2H, t, J = 3.0Hz, 2-CH2), 4.15(1H, s, 5-CH) 6.79-8.12(14H, m, ArH). 13C NMR (CDCl3, δ ppm): 34.54(CH2S), 45.43(CH2), 46.72(CH), 88.23(C-N), 127.32-126.54(C = C), 146.41-132.32(Caro), 183.54(C = O). EI-MS: m/z (%) = 413(M+ 32), 385(39), 235(100), 108(35). Anal. Calcd for C25H19NOS2: C, 72.61%; H, 4.63%; N, 3.39%. Found: C, 72.73%; H, 4.78%; N, 3.86%.

(5R,8R)-spiro-{6-ethoxycarbonyl-7-phenyl-1-aza-4-thia-bicyclo [3,3,0]-6-octene-8, 3'}-benzo[b]thiophene-2'-one (9)

Dark brown power, yield: 0.31g (69%), mp: 110-112°C. IR (KBr): 1710(C = O), 1410(C-N), 690(C-S) cm-1. 1H NMR (CDCl3, δ ppm): 1.25(3H, t, J = 2.4Hz Me), 2.61(2H, t, J = 2.9Hz 3-CH2), 2.69(2H, t, J = 3.1Hz, 2-CH2), 2.67(2H, t, CH2), 3.75(2H, q, J = 6.3Hz, OCH2),4.12 (1H. s,5-CH) 7.49-7.94(4H, m, ArH). 13C NMR (CDCl3, δ ppm): 33.85(CH2), 44.96(CH), 46.32(CH2), 84.36(C-N), 124.32-122.93(C = C), 140.14-130.12(CHaro), 180.25(O = C-O), 185.44(C = O). EI-MS: m/z (%) = 321(M+ 38), 275(100), 261(20), 284(15), 234(18). Anal. Calcd for C22H19NO3S2: C, 64.52%; H, 4.68%; N, 3.42%. Found: C, 64.68%; H, 4.79%; N, 3.54%.

Declarations

Acknowledgements

Financial assistance from UGC New Delhi (Project No.36-289/2008(SR)) is gratefully acknowledged. Sonali Verma also thanks UGC for RGN-SRF.

Authors’ Affiliations

(1)
Department of Chemistry, University of Rajasthan

References

  1. Katritzky AR, Ramsden CA: Comprehensive Heterocyclic Chemistry III. Edited by: Scriver EFV, Taylors RJK. Elsevier, New York; 2008.Google Scholar
  2. Harwood LM, Vickers RJ: In synthetic applications of 1,3-dipolar cycloadditions chemistry toward heterocycles and natural products. Edited by: Padwa A, Pearson WH. Wiley, New York; 2003:169.Google Scholar
  3. Gothelf KV, Kobayashi S, Jorgensen KA, eds: In cycloaddition reactions in organic synthesis. Wiley-VCH, Weinheim, Germany; 2002:211.
  4. Zang H, Chan WH, Lee AWM, Xia P-F, Wong WY: Asymmetric 1,3-dipolar cycloaddition of chiral α, β-unsaturated-γ-sultams with nitrile oxides and nitrones. Lett Org Chem 2004, 1: 63–66. 10.2174/1570178043488644View ArticleGoogle Scholar
  5. Najera C, Sansano JM: 1,3-Dipolar cycloaddition reactions of metal stabilized azomethine ylides. Org Biomol Chem 2009, 7: 4567–4581. 10.1039/b913066gView ArticleGoogle Scholar
  6. Pandey G, Banerjee P, Gadre SR: Asymmetric 1,3-dipolar cycloaddition reactions of metal stabilized azomethine ylides. Chem Rev 2006, 106: 4484–4517. 10.1021/cr050011gView ArticleGoogle Scholar
  7. Girgis AS: Regioselective synthesis of dispiro[1H-indene-2, 3'- pyrrolidine-2', 3''-[3H]indole]-1, 2''[1''H]-diones of potential anti-tumour properties. Euro J Chem 2009, 44: 91–100.View ArticleGoogle Scholar
  8. Tsuge O, Kanesmasa S: In Advances in Cycloaddition. Volume 3. Edited by: Curran DP. Jai Press, Greenwich, CT; 1993:99.Google Scholar
  9. Poornachandran M, Jayagobi M, Ragunathan R: Synthesis of novel spiropyrrolizidines as potent antimicrobial agents for human and plant pathogens. J Chem Res 2009, 4: 240–243.View ArticleGoogle Scholar
  10. Liang Y, Song YZX: Phosphine- and water- cocatalyzed [3 + 2] cycloaddition reactions of 2-Methyl-2,3-butadienoate with fumarates: A computational and experimental study. Synlett 2009, 6: 905–909.Google Scholar
  11. Rajopadhye M, Popp FD: Chemistry of benzo[b]thiophene-2,3- dione. Heterocycles 1988, 27: 1489–1502. 10.3987/REV-87-382View ArticleGoogle Scholar
  12. Navarro A, Sanchez Pino M-J, Gomez C, Bandez MJ, Cadenas E, Boveris A: Dietary thioproline decreases spontaneous food intake and increases survival and increases neurological function in mice. Antioxid Redox Signal 2007, 9: 131–141. 10.1089/ars.2007.9.131View ArticleGoogle Scholar
  13. Li C, Han BY, Wanshun L: Antioxidant activity of N-acetyl glucosamine based thiazolidine derivative. High Tech Lett 2007, 13: 441–445.Google Scholar
  14. Padwa A: 1,3-Dipolar Cycloaddition Chemistry. Wiley, New York; 1984.Google Scholar
  15. Arora K, Jose D, Singh D, Gupta RS, Pardasani P, Pardasani RT: Stereoselective synthesis and antioxidant activity of azabicycloadducts derived from9, 10-phenanthrenequinone. Heteroatom Chem 2009, 20: 379–392.Google Scholar
  16. Frisch MJ, et al.: GAUSSIAN 03( Revision B.03 ). Gaussian, Inc., Wallingford, CT; 2004.Google Scholar
  17. Amornraksa K, Grigg R, Gunaratne HQN, Kemp J, Sridharan V: X:YZH Systems as potential 1, 3-dipoles. Part 8. Pyrrolidines and Δ 5 -pyrrolines(3, 7-diazabicyclo[3.3.0]octenes) from the reaction of imines of α-amino acids and their esters with cyclic dipolarophiles. Mechanism of racemisation of α-amino acids and their esters in the presence of aldehydes. J Chem Soc Perkin Trans 1987, I: 2285–2296.View ArticleGoogle Scholar
  18. Perrin DD, Armarego WLF, Perrin DR: Purification of Laboratory Chemicals. 2nd edition. Pergamon Press, Oxford; 1998.Google Scholar
  19. Vogel AI: Vogel's Text Book of Practical Organic Chemistry. 4th edition. ELBS Longman, London; 1984.Google Scholar

Copyright

© Verma et al; licensee Springer. 2011

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.