7th International Electronic Conference on Synthetic Organic Chemistry (ECSOC-7), http://www.mdpi.net/ecsoc-7, 1-30 November 2003

[A011]

 

 

Cycloaddition coupling reactions of ethyl-5,6-diphenyl-1,2,4-triazine-3-carboxylate with norbornenes and 7-oxanorbornenes.

 

 

Davor Margetić,*A Ronald N. WarrenerB and Douglas N. ButlerB

 

A Laboratory for Physical Organic Chemistry,

Department of Organic Chemistry and Biochemistry,

Ruđer Bošković Institute, Bijenička c. 54, 10000 Zagreb, Croatia

(*) Corresponding author. Email: margetid@emma.irb.hr
B Centre for Molecular Architecture,

Central Queensland University, Rockhampton, Queensland, 4702, Australia

____________________________________________________________________________________________________________________

Abstract. In this paper, reactions of substituted 1,2,4-triazines with norbornenes and 7-oxanorbornenes are described. Depending on the substrate and the reaction conditions, 1:1 or 2:1 deazetized cycloaddition products were obtained. Furthermore, a novel aza-isobenzofuran was prepared as an intermediate by flash vacuum pyrolysis of one on these adducts.

____________________________________________________________________________________________________________________

 

Introduction.  The convergent BLOCK building approach[1] to the synthesis of rigid polyalicyclic systems offers great advantages over stepwise sequential cycloaddition protocols.[2,3,4] This synthetic approach allows efficient, stereoselective coupling of small norbornene fragments into very complex structures and the continuing development of molecular “glue” reagents based on tandem Diels-Alder reactions is an ongoing project in our research group. Cyclic dienes such as 2,5-bis(trifluoromethyl)-1,2,4-oxadiazole and 2-pyridyl-s-tetrazine have been commonly used as coupling agents in for these reactions. Our preliminary work with 1,2,4-triazine 2b[5] revealed that this molecule could also be used as a “glue” to fuse norbornenes together.[5] Thus triazine 2b was used to deliver the phenanthroline chromophore to a polynorbornane scaffold in the process of linking two norbornenes. In the work presented here, triazines 2a-2c[6] were used to establish their relative cycloaddition reactivity and to further develop the process of using triazines as “glue” reagents. We also report the synthesis of a novel azaisobenzofuran (aza-IBF) as a trappable intermediate.

 

Results and discussion.

 

Part I. Cycloaddition Experiments

 

We have found that triazines 2a and 2b react very poorly with the olefinic site of the exo-furan/maleimide adduct 1. Only traces of products were detected by proton NMR regardless of the reaction conditions employed (thermal - sealed glass tube, overnight in chloroform, or at high pressure - at 10 kbar, 3 days, DCM, RT) (Scheme 1). In contrast, it was found that ethyl-5,6-diphenyl-1,2,4-triazine-carboxylate 2c was more reactive and yielded a number of products on reaction with 1. The product distribution depended on whether the reaction was conducted under thermal or photochemical conditions.

 

 

Under thermal conditions, compound 4 was the only product obtained (at 140 oC, sealed tube, overnight, 16 % yield) from reaction of 1 with triazine 2c and it was obviously formed by aromatisation of the initially formed exo,endo- product 3. We have not attempted to optimize yields for reported reactions. It is not yet clear whether this 3 ® 4 oxidation was facilitated by the presence of triazine or by air oxidation. Experiments conducted in a sealed tube suggested that the oxidation by triazine was a more likely explanation. In a separate thermal experiment, triazine 2c was reacted with alkene 1, followed directly by DDQ oxidation to afford product 4 in 27 % yield. Under high-pressure conditions, product 4 was not isolated but products 3 and 5 (in ~1:1 ratio, and 18 % and 11 % isolated yield, respectively) were obtained. Exo,endo-  product 3 is the intermediate formed by addition of triazine 2c to alkene 1, followed by dinitrogen extrusion. Regarding the formation of adduct 5, it is interesting to note that this oxanorbornene coupling reaction occurs stereoselectively, yielding only the bent isomer exo,endo- 5 with no evidence for the formation of the linear adduct  exo,exo- 6 being observed. This is in a sharp contrast with our experience with the 1,2,4-triazine coupling of norbornenes or cyclobutenes reported by us earlier as well as the previously prepared s-tetrazine coupled product 7[5] where only the exo,exo-product was isolated.[1,7,8] We have tried the high-pressure reaction of triazine 2c with similar 7-oxa-bridged substrate 8 (Figure 1) (14 kbar, 7 days, DCM, room temperature), but the reaction mixture was too complex to be able to draw any conclusions with the attempted separation of this reaction mixture not yielding any discrete products. Another standard alkene 9 possessing a 7-oxa-bridge, viz., 7-oxabenzonorbornadiene was not used, since it would be expected to certainly yield isobenzofuran as a fragmentation product.[9]

 

 

The structure of product 5 was established on the basis of the proton-NMR spectroscopic data with its significant lack of molecular symmetry (Figure 2). Two N-methyl signals occurred at d 2.85 and 2.88 ppm confirming that the N-methylsuccinimide moieties were in different environments. There are four oxa- bridghead protons and four doublets assignable to the two coupled endo- and the two coupled exo- protons.

 

Additional supporting evidence of this structural assignment was obtained from 2D- COSY and NOESY experiments. The most indicative is the absence of correlation between endo- protons H5 H6 and exo- protons H7 H8 in NOESY spectrum. It shows that molecule 5 has bent structure. Furthermore, COSY and NOESY experiments indicate that there are two separated groups of protons, belonging to two 7-oxanorbornane moieties (H1-H6 and H7-H12), without any correlation between them.

 

Reaction of triazine 2b and dimethoxynaphthonorbornadiene 10 (chloroform, 6h, 160 oC, or 10 kbar, DCM, 3 days) gave no isolable product. However, when triazine 2c was reacted with the same alkene 10 under various conditions, four different products 11-14 were identified (separated by radial chromatography) from these reaction mixtures (Scheme 2). Three of these were anticipated products arising from initial 1:1 cycloaddition while the fourth was the derived 2:1 product 14. These experiments were conducted using equimolar amounts of triazine and alkene and it is important to note that the lower temperature reactions (chloroform, sealed tube, 6h, 160 oC, or 10 kbar, DCM, 3 days, RT) gave products 11[11], 12 and 14 (in 6 %, 9 % and 5 % yields, respectively), while a higher temperature thermal coupling experiment at 180 oC (overnight) gave products 12, 13 and 14 (in 8%, 6% and 9% yields, respectively). This was the only experiment where product 13 was observed. In the proton NMR spectrum of product 13, there was a characteristic singlet at d 8.54 which was assigned to the NH proton. Furthermore, a thermal reaction conducted at 140 oC (overnight, large excess of alkene) gave a mixture of products 12 and 14 in 1:6 ratio, while a similar experiment at 100 oC (overnight) gave a mixture of 12 and 14 in 2:1 ratio. In the proton NMR spectrum of the 1:2 product 14, the most important signals for structural assignment were the methylene protons Ha and Hb which were seen at d 2.31 and 1.36, respectively. Due to the steric compression[11] of  the methylene bridge caused by the C=N bridge, proton Hb was significantly shifted upfield. All assignements were supported by COSY and NOESY correlations.

 

 

These values are surprisingly at lower magnetic field than observed for structurally related triazine and tetrazine products 15-17 (shown in Figure 3) and may be rationalised by the magnetic influence of phenyl or pyridyl substituents. A more comparable example is found in the a-dione adduct 18.[12]

 

 

Finally, thermal reactions of alkene 10 with triazine 2a (conducted at temperatures of 140-180 oC) gave only product 19 in 25 % yield (Scheme 3). There was no reaction observed under standard high pressure conditions (10 kbar, 3 days, DCM). A small amount of HCl adduct 20 was also isolated in a thermal experiment conducted at 140 oC (7 % yield). It was assumed that the solvent was the HCl source to give this unexpected addition product.

 

 

 

Part II. Flash Vacuum Pyrolysis Experiments

 

Some flash vacuum pyrolysis (FVP) experiments were conducted to explore the preparation of the novel aza-izobenzofuran (aza-IBF) system and to study its chemical reactivity. Retro-Diels Alder (RDA) methodology was chosen since there were many examples of its successful use in the preparation of various IBFs. [13] Thus, FVP of exo- product 4 (at 500 oC and 0.005 mbar) gave the endo-adduct 22 (Scheme 4) (isolated in 4% yield). We assume that isobenzofuran 21[14] was formed at the first step and that this reacted immediately with the N-methylmaleimide which formed by the pyrolysis and deposited on the cooler walls of the FVP tube outside the tube furnace. If only the exo-compound 4 had been isolated, it would be not possible to tell if retro- Diels-Alder reaction had occured or not, but formation of the endo- adduct was taken as very supporting evidence. In order to optimize this retro reaction, FVP experiments at various temperatures were conducted (470, 500 and 550 oC, all at 0.005 mbar). At 470 oC mainly unchanged substrate was detected, with some other minor products, but at 500 oC and 550 oC, similar reaction mixtures were obtained, without any real improvement in the amount of 21 being formed. It was interesting to note that when the crude products were washed from the pyrolysis tube with chloroform, a bright green fluorescent solution was obtained. There was also an amount of a white polymeric material left on the tube walls that was not soluble in either chloroform, acetone or methanol and we propose that this material was an aza-IBF polymer.

 

 

When the above fluorescent chloroform solution was treated with N-methyl maleimide, the fluorescent color immediately disappeared suggesting that that the fluorescent material was the aza-IBF 21. A chromatographic separation of the crude FVP mixture gave a colored, fluorescent material in the first fraction that might have been the aza-IBF in a more pure form. From this fluorescent fraction, we obtained some NMR evidence of aza-IBF formation, viz., singlets at d 7.99 (1H) and 8.75 (1H) that were assigned to the two furan signals. These ‘aromatic’ lines are at the expected positions for IBF signals, which are usually shifted downfield as compared to the furan signals.[15] Furthermore, methylene protons of the ethyl group occur as an AB pair of anisochronous protons at d 4.57 and 4.53 ppm (J=6.1 Hz). The same chromatographic separation also gave product 23 (in 17 % yield), the dehydration product of 4 (or 22).

 

Conclusion.

The role of 1,2,4-triazines to act as a molecular glue have been confirmed, however the reaction is not particularly efficient.  Oxidation of the dihydropyridine product formed by deazetisation of the first-formed 1:1-adduct, occurred much more readily than that observed with the corresponding dihydropyridazines formed in the corresponding 3,6-di(2-pyridyl)-tetrazine reaction.  Evidence for the formation of a fluorescent azaisobenzofuran was supported by trapping with maleimides, but clearly its stability is low as evidenced by rapid polymer formation during isolation from the FVP experiments.

 

Acknowledgements.  We are grateful to the Australian Research Council (ARC) for funding.

 

References. 

 

1. Warrener, R. M.; Butler, D. N.; Russell, R. A. Syntlett 1998, 566.

2. Lawson, J. M.; Craig, D. C.; Oliver, A. M.; Paddon-Row, M. N. Tetrahedron 1993, 51, 3841.

3. Kumar, K.; Tepper, R. J.; Zeng, Y.; Zimmt, M. B. J. Org. Chem. 1995, 60, 4051.

4. Lin. C.-T.; Hsu, H.-C.; Wang, M.-F.; Chou, T.-C. Tetrahedron 2000, 56, 5383.

5. Warrener, R. N.; Margetic, D.; Russell, R. A., Synlett 1998, 585.

6. Elix, J. A.; Wilson, W. S.; Warrener, R. N.; Calder, I. C. Aust. J. Chem. 1972,  25(4), 865.

7. Margetic, D.; Warrener, R. N. unpublished results, 1997.

8. Margetic, D.; Johnston, M. R.; Warrener, R. N., Third International Electronic Conference on Synthetic Organic Chemistry (ECSOC-3), http://www.reprints.net/ecsoc-3.htm, September 1-30, 1999, Editors: Esteban Pombo-Villar, Reinhard Neier and Shu-Kun Lin, CD-ROM edition ISBN 3-906980-04-9 Published in 2000 by MDPI, Basel, Switzerland.

9. Warrener, R. N. J. Am. Chem. Soc. 1971, 93, 2346.

10. Similarly, thermally conducted reaction of norbornene with 2c gave only 1:1 aromatized adduct of type 12.

11. Margetic, D.; Johnston, M. R.; Warrener, R. N.; Butler, D. N., Article 37, The Fifth International Electronic Conference on Synthetic Organic Chemistry (ECSOC-5), http://www.mdpi.org/ecsoc-5.htm, September 1-30, 2001, Editors: Oliver Kappe, Pedro Merino, Andreas Marzinzik, Helma Wennemers, Thomas Wirth, Jean-Jacques Vanden Eynde and Shu-Kun Lin, CD-ROM edition ISBN 3-906980-06-5 Published in 2001 by MDPI, Basel, Switzerland.

12. Johnston, M. R.; Latter, M. J.; Warrener, R. N., Golic, M.; McKavanagh, D.; Margetic, D., Fourth International Electronic Conference on Synthetic Organic Chemistry (ECSOC-4), Paper No: A0057, http://www.unibas.ch/ mdpi/ecsoc-4.htm, September 1-30. 2000, Editors: Thomas Wirth, C. Oliver Kappe, Eduard Felder, Ulf Diederichsen and Shu-Kun Lin, CD-ROM edition ISBN 3-906980-05-7 Published in 2000 by MDPI, Basel, Switzerland.

13. Wiersum, U. E. Rec.Trav. Chim. Pays-Bas 1982, 102, 365.

14. Ethyl-5,6-diphenyl-furo[3,4-c]pyridine-2-carboxylate; for preparation of parent unsubstituted compound see: Wege, D. in "Advances in Theoretically Interesting Molecules", 1998, Vol. 4, 1, R. P. Thummel Ed, JAI Press, Stamford, Connecticut and London, England; Wiersum, U. E.; Elred, C. D.; Vrijhof, P.; van der Plas, H. Tetrahedron Lett. 1977, 1741; substituted derivatives: Sarkar, T. K.; Ghosh, S. K.; Nandy, S. K.; Chow, T. J.; Tetrahedron Lett. 1999, 40, 397; Muller, P.; Schaller, J.-P. Tetrahedron Lett. 1989, 30, 1507; Chen. C.-W.; Beak, P. J. Org. Chem. 1986, 51, 3325.

15. Stringer, M. B.; Wege, D. Tetrahedron Lett. 1980, 21, 3831.