Formation of oligotriazoles catalysed by cucurbituril

Abstract

The catalytic activity of cucurbituril in 1,3-dipolar cycloadditions has been applied to the synthesis of oligotriazoles.

---

The supramolecular chemistry of cucurbituril is unique amongst cavitands due to its multiple modes of molecular recognition.^1,2 Cucurbituril binds strongly alkali and alkali earth metal cations and ammonium ions.^1–3 It encapsulates aliphatic^4–7 and aromatic molecules^3,8,9 most of which are substituted with mono- or di-ammonium ions required for strong binding. Probably its most remarkable feature is the ability to catalyse 1,3-dipolar cycloadditions, between suitably substituted aliphatic azides and terminal alkynes in a regioselective fashion.^1,10,11 Mock et al.¹⁰ were the first to study the catalytic mechanism of this reaction and after detailed investigations¹¹ concluded that rate acceleration to the triazole product is a consequence of release of strain present in the ternary complex involving cucurbituril, an alkyne and an azide. More recently a very similar catalytic process was designed by Sanders et al.¹² through the elegant and careful construction of cyclic metalloporphyrin trimers which were shown to catalyse a Diels–Alder reaction with regiostereochemical control inside their cavities.

Not unexpectedly Mock et al. went on to explore cucurbituril as a catalyst for the polyaddition reaction between dialkyne and diazide monomers but without success.¹ Product inhibition was claimed to be the reason for the very low catalytic activity of cucurbituril since only a mixture of mono- and di-triazoles had formed.

Since then polymers involving cucurbituril have been synthesised but in all cases cucurbituril has been part of a directed polyrotaxane synthesis making use of preformed pseudorotaxane monomers.^6,13,14 As a catalyst for polymer synthesis cucurbituril has only been revisited once by Steinke et al.¹⁵ who used the cavitand in stoichiometric amounts (thereby avoiding product inhibition) which resulted in the successful synthesis of mainchain polyrotaxanes.

After our initial attempts to prepare cucurbituril-catalysed polytriazoles with azidoethyl and propargyl disubstituted hexanediamines had been unsuccessful¹⁵ we decided to repeat Mock’s original experiments^10,11 aiming at improving our monomer design in the process (Fig. 1).

Fig. 1 Synthesis of oligotriazole pseudooligorotaxanes 4 and free oligotriazoles 5. (i) CH=CCH₂NH₂, (BOC)₂O; DMAP, dioxane, 20 °C, 48 h, 92%; (ii) CH=CCH₂Br, DMF, NaH, then Et₂O, HCl_g, 80%; (iii) bis(2-chloroethyl)amine HCl, NaN₃, H₂O, 6 h rt, 2 h 70 °C, 60%; (iv) 0.2 M Na₂SO₄ in D₂O (see Table 1 for details), (v) 1 M NaOH, pH 12, CHCl₃ extraction (see Table 1 for details).

Bis(propargyl)amine HCl (1) was synthesised by reacting BOC protected propargylamine with propargyl bromide followed by treatment with HCl. Bis(2-aziodethyl)amine HCl (2) was accessed through bis(chloroethyl)amine HCl and cucurbituril (3) was obtained using the protocol by Kim et al.¹⁶ Reactions were carried out in 0.2 M Na₂SO₄ in D₂O or DCl in D₂O (20% wt) using equimolar ratios of monomers (1) and (2) and concentrations of cucurbituril ranging from 2 to 0.1 equivalents. Reaction conditions are summarised in Table 1. The rotaxinated oligotriazole backbone (4) was separated from cucurbituril afterwards to obtain free oligotriazole (5) by extracting the basified aqueous solution with chloroform (Table 1).

Table 1 Reaction conditions and ¹H NMR data for PT1–PT5

	Molar equiv.
				Time/temp.
Compound	[1]	[2]	[3]	(h/°C)	[Triazole]/[3]
a All reactions were carried out in 0.2 M Na2SO4 in D2O apart from PT1 where DCl–D2O (20% wt) was used instead.
PT1-A ^a	1	1	2	168/25	0.4
PT2	1	1	2	168/25	0.7
PT3-A	1	1	0.5	168/25	2.0
PT3-B	1	1	0.5	336/25	2.1
PT4-A	1	1	0.1	168/25	1.4
PT5	1	1	0	336/25	0.0

---

Analysis of the ¹H NMR spectra of reactions PT1 to PT5 focused on the ratio of the integration of the newly formed triazole ring protons to cucurbituril protons (Table 1). PT5 is the ‘blank’ experiment and it was found that no triazole rings are formed in the absence of cucurbituril under otherwise identical reaction conditions. For reactions PT1 to PT4 however triazole formation was observed. If more than one triazole ring is formed for every cucurbituril macrocycle (3) present ([triazole]/[3] > 1; Table 1) it means that on average cucurbituril has taken part in more than one catalysed 1,3-dipolar cycloaddition indicative of catalytic turnover. This is the case for reactions PT3-B and PT4-A in which subequimolar amounts of cucurbituril to monomer have been used and after 336 h at rt the maximum number of triazole rings had been formed. Subsequent heating of the samples at 80 °C for 96 h induces further triazole formation. In these cases the cycloaddition reaction is most likely induced thermally and not a result of the presence of cucurbituril. Equimolar amounts of cucurbituril to azide and alkyne functional groups present, as for examples PT1 and PT2, led to ratios of triazole to cucurbituril of less than one. This was not unexpected for PT1 since the reaction was carried out at low pH similar to the work by Mock et al. (aqueous formic acid (H₂O–HCOOH (1∶1)) and the result is consistent with their findings.^11,17 The neutral aqueous 0.2 M Na₂SO₄ medium for PT2 was found to lead to higher conversion than the acidic reaction medium (PT1). We associate the increase in the formation of triazole rings (PT2) with the difference in pH.¹⁸ Dissociation is accelerated at higher pH and thus leads to an increase in the observed rate of the dissociation as shown by Mock et al. using model compounds.¹⁹ The further increase of the ratio of triazole to 3 when reducing the amount of 3 to subequimolar levels (PT3 and PT4) could be an indication of the rate of dissociation being affected by an associative displacement mechanism (the presence of an excess of a guest accelerates dissociation of the substrate encapsulated by 3) also experimentally verified by Mock et al.¹⁹ However based on our studies on the threading behaviour of 3 onto a poly(hexamethylene imminium) backbone,²⁰ we expect at high concentrations of 3 path c to become neglible and path b to be dominant because of the high association constant and low rate of dissociation of the monotriazole pseudorotaxane (Fig. 2). Association of a second molecule of cucurbituril adjacent to the first through sharing of the ammonium ion of monomer fragment 2 is sterically unfavourable²⁰ and the correct conformation of the ternary complex required for the catalysis is not accessible.¹¹ At subequimolar concentrations of 3 on the other hand path b is no longer dominant and catalytic turnover becomes possible via path c.

Fig. 2 Catalytic pathways of oligotriazole formation.

Table 2 shows representative MS data (FAB) for reactions PT1 and PT3. In both cases we find evidence for the formation of oligotriazoles in which up to 9 monomers (1 and 2; PT1) have been joined to form an oligotriazole chain containing the same number of triazole rings. For PT1 we have also included data showing the composition of the organic phase after extraction of the aqueous phase with chloroform, also confirming the presence of oligotriazole chains. All of the data are in full agreement with the findings of our NMR analysis demonstrating the ability of cucurbituril to exhibit catalytic turnover albeit very sluggishly. We interpret the presence of higher triazole oligomers as the result of different catalytic pathways being available simultaneously thus resulting in a distribution of oligotriazole chain lengths. Fig. 2 captures the various pathways potentially involved in the overall catalytic process.

Table 2 FAB MS data for PT1 and PT3

PT3-B	MW (theo.)	MW (det.)	PT1	MW (theo.)	MW (det.)	PT1 extr.	MW (theo.)	MW (det.)
[1] and [2] = monomers, [3] = cucurbituril (Fig. 1), det. = detected, extr. = extracted.
[1]2[2]2	640.6	644.1	[1] + [3] − Cl	1090.2	1087.7	[1][2]	248.3	249.1
[1]3[2]2 + 2H2O − N2	772.7	771.8	[1]2[2] + [3] + H2O	1463.4	1463.2	[1]2[2]2	496.6	497.3
[1]3[2]3 + 2H2O − N2	962.9	964.4	[1][2]2 + [3] − Cl	1490.5	1491.2	[1]3[2]3	744.9	744.5
[1][2] + [3] + 2H2O	1352.3	1350.2	[1]3[2]3 + [3] 3H2O	2010.9	2014.3	[1]4[2]3	838.0	835.8
[1]5[2]4 − Cl	1375.3	1374.4	[1]5[2]4 + [3] + H2O − N2	2397.2	2395.5

---

After the formation of the first triazole ring (patha) a second one is formed by either following path b (followed by d) or path c (followed by e). Path c and e are required for catalytic turnover whereas path b illustrates reduction (or suppression) of catalytic activity caused by steric crowding and high affinity of the catalyst for the product.

Further investigations into the role of the solvent and the stoichiometry between 3 and each monomer are currently underway to establish further mechanistic details of the catalytic process with the aim to prepare polytriazoles in this fashion.

We would like to thank Mr Christos and Mrs Froso Krasia for financial support and Mr John Barton for the mass spectrometry data. We would also like to thank one of the reviewers for helpful comments.

Notes and references

1. W. L. Mock, Top. Curr. Chem., 1995, 175, 1.
2. P. Cintas, J. Incl. Phenom. Mol. Recognit. Chem., 1994, 17, 205.
3. S. Karcher, A. Kornmuller and M. Jekel, Acta Hydrochim. Hydrobiol., 1999, 27, 38.
4. W. L. Mock and N. Y. Shih, J. Am. Chem. Soc., 1988, 110, 4706.
5. C. Meschke, H. J. Buschmann and E. Schollmeyer, Thermochimica Acta, 1997, 297, 43.
6. K. M. Park, J. Heo, S. G. Roh, Y. M. Jeon, D. Whang and K. Kim, Mol. Cryst. Liq. Cryst. Sci. Technol., Sect. A, 1999, 327, 65.
7. H. J. Buschmann, K. Jansen and E. Schollmeyer, Thermochim. Acta, 2000, 346, 33.
8. R. Neugebauer and W. Knoche, J. Chem. Soc., Perkin Trans. 2, 1998, 529.
9. H. J. Buschmann and E. Schollmeyer, J. Incl. Phenom. Mol. Recognit. Chem., 1992, 14, 91.
10. W. L. Mock, T. A. Irra, J. P. Wepsiec and T. L. Manimaran, J. Org. Chem., 1983, 48, 3619.
11. W. L. Mock, A. Irra, J. P. Websiec and M. Adhya, J. Org. Chem., 1989, 54, 5302.
12. M. Marty, Z. Clyde-Watson, L. J. Twyman, M. Nakash and J. K. M. Sanders, Chem. Commun., 1998, 2265.
13. C. Meschke, H. J. Buschmann and E. Schollmeyer, Polymer, 1999, 40, 945.
14. E. Lee, J. Heo and K. Kim, Angew. Chem., Int. Ed., 2000, 39, 2699.
15. D. Tuncel and J. H. G. Steinke, Chem. Commun., 1999, 1509.
16. J. Kim, I. S. Jung, S. Y. Kim, E. Lee, J. K. Kang, S. Sakamoto, K. Yamaguchi and K. Kim, J. Am. Chem. Soc., 2000, 122, 540.
17. W. L. Mock, T. A. Irra, J. P. Wepsiec and M. Adhya, J. Org. Chem., 1989, 54, 5302.
18. Remarkable solvent effects for cucurbituril are already known for the encapsulation of aliphatic mono- and di-ammonium compounds with small changes in solvent polarity impacting significantly on complexation constants for encapsulated protonated aliphatic amines (ref. 5).
19. W. L. Mock and N. Y. Shih, J. Org. Chem., 1986, 51, 4440.
20. D. Tuncel and J. H. G. Steinke, Chem. Commun., 2001, 253.

---

Footnote

† Current address: Max-Planck-Institute for Colloids and Interfaces, Golm, 14424 Potsdam, Germany.

---