Probing cucurbit[8]uril-mediated supramolecular block copolymer assembly in water using diffusion NMR

Abstract

Diffusion NMR and solution viscometry were used to probe the cucurbit[8]uril-mediated host–guest self assembly of multiple molecular guests to form 5-component supramolecular ABA triblock copolymers in aqueous solution.

---

Cucurbit[8]uril (CB[8]) is one of the larger homologues of the macrocyclic, barrel-shaped cucurbituril family of molecular hosts.^1,2 Its hydrophobic cavity is large enough to accommodate two molecular guests, forming ternary complexes in water.³ These ternary complexes usually consist of an electron-deficient viologen (paraquat) first guest and an electron-rich aromatic second guest such as naphthol,⁴ associating with the CB[8]-host sequentially in a stepwise manner.³ The complexes are stable in water (K_a = 10^8–12 M⁻²)^4,5 as well as stimuli-responsive.⁶ Recently, CB[8]'s complex formation with polymeric guests to form supramolecular copolymers has been demonstrated both in solution⁷ and in the gas phase.⁸ Supramolecular polymers^7,9–13 exhibit intelligent self-assembly akin to that seen with biological building blocks,¹⁴ and have attracted much interest, demonstrating a range of stimuli-responsive properties both in solution¹⁵ and in the bulk.^16,17 Supramolecular polymers therefore combine the material features of macromolecules, such as the ability to form gels,¹⁸ and higher-ordered, compartmentalised structures,^19,20 with the tunability offered by assembling individual small-molecule building blocks.

Previously, our group has demonstrated the CB[8]-mediated assembly of supramolecular AB diblock copolymers from viologen- and naphthol-terminated polymers based on the hydrophilic poly(ethylene glycol) (PEG), and hydrophobic poly(isoprene).⁷ In the presence of CB[8], a naphthol guest-terminated PEG, A block, and a low molecular weight ditopic viologen guest, B block (MVdimer), were found to form discrete supramolecular ABA triblock copolymer complexes.⁸ The discrete, multi-component assemblies remained stable even in the gas phase as probed by nano-electrospray mass spectrometry (nano-ESI-MS).⁸ In tandem with ongoing studies on the self-assembly of CB[8]-complementary polymers in water, we sought to apply the versatility of diffusion-ordered (DOSY) NMR²¹ towards probing the CB[8]-mediated ABA triblock copolymer formation in solution. This non-invasive, and in situ technique maps the assembly process as a 2-dimensional NMR spectrum, correlating chemical shift against diffusion coefficient (D).^22,23

Fig. 1 shows the components used in this study; a triethylene glycol (TEG)-spaced viologen dimer MVdimer, and two polymers, a naphthol-terminated PEG (P1, M_n = 1100–8000 g mol⁻¹) and a dibenzofuran-terminated poly(N-isopropylacrylamide) (PNIPAAM) (P2, M_n = 19,000 g mol⁻¹). The dibenzofuran polymer end-group is used here simply as a suitable alternative second guest to naphthol.⁴

Fig. 1 The synthetic building blocks used in this self-assembly study, P1 (M_n = 1100–8000 g mol⁻¹), P2 (M_n = 19,000 g mol⁻¹), MVdimer, and CB[8].

A pulsed field-gradient stimulated-echo (PFGSE) sequence with 3-9-19 water suppression (stebpgp1s19)²⁴ was used for all DOSY measurements. Fig. 2 shows a 2D DOSY plot of MVdimer and P1_5K (M_n = 5000 g mol⁻¹) prior to the addition of CB[8] (Fig. 2, top). The diffusion coefficients of the two components are clearly separated as the polymeric and small molecule guests diffuse as single molecular species (log D = −10.12 and −9.53 m² s⁻¹ respectively). Upon adding CB[8] (Fig. 2, bottom), the NMR signals of both MVdimer and P1_5K now share a single diffusion coefficient (log D = −10.2 m² s⁻¹), diffusing as one entity. The aromatic signals of MVdimer and P1_5K also showed upfield shifts and significant line-broadening by NMR upon complexation as is typical for CB[8]-binding.^7,8

Fig. 2 A 2-dimensional water-suppressed DOSY plot illustrating a two-component system consisting of P1_5K and MVdimer prior to adding CB[8] (top), and a five-component ternary complex after CB[8] addition (bottom). 500 MHz, D₂O at 20 °C.

A 1-dimensional Stejskal-Tanner plot^25,26 (Fig. 3) of decreasing signal intensity (I/I₀) over increasing gradient strength (b parameter, Fig. 3, caption) shows the acquired diffusion data instead as a straight line graph where the slope equals the diffusion coefficient. In Fig. 3 a very significant decrease in diffusion of MVdimer is observed upon adding CB[8] to a solution of the high molecular weight P2 (M_n = 19,000 g mol⁻¹) and MVdimer (2 : 1). The MVdimer then diffuses with the same diffusion coefficient as P2 upon ternary complex formation.

Fig. 3 A Stejskal-Tanner plot showing a two-component system consisting of P2 (M_n = 19,000 g mol⁻¹) and MVdimer prior to adding CB[8], and the five-component ternary complex after CB[8] addition. The change in D_MVdimer upon CB[8] addition is especially marked (curled arrow). Slopes (log D/m² s⁻¹) are shown in brackets. Key: MVdimer_free (blue), P2_free (green), MVdimer_bound (yellow) and P2_bound (red). 500 MHz, D₂O at 20 °C. b = 4π² ·γ² · G_i² ·δ² ·(Δ − δ/3)·1 × 10⁻⁵.

Both complexes with P1 and P2 were shown to be discrete, 5-component (2 : 1 : 2) complexes both in the gas phase (nano-ESI-MS) and in solution;⁸ In the case of P1-based complexes, ITC (Table S1 and Fig. S4–7, ESI†) confirmed this for a number of molecular weights (M_n = 1100–8000 g mol⁻¹). An MVdimer-CB[8] (1 : 2) binary complex forms initially, which then incorporates two molecules of polymeric guest (P1, P2) to form the 5-component ternary complex.³ To probe the molecular weights of the complexes further, a calibration plot (Fig. 4) was prepared relating the diffusion coefficients of commercially available hydroxy-terminated PEGs (HO-PEG) to their molecular weight (M_n = 200–20,000 g mol⁻¹, blue circles). Overlayed on this plot are P1_free (M_n = 1100–8000 g mol⁻¹, green circles) along with the ternary complexes formed on adding CB[8] (hollow red circles). All solutions of HO-PEG and P1 were measured in the presence of MVdimer (2 : 1 stoichiometry) so as to keep the solution viscosity constant between samples. The complexes were predicted to approximately double in weight upon CB[8] addition when compared to their respective P1_free based on a 2 : 1 : 2 stoichiometry. A good overlap was observed as shown in Fig. 4, and the doubling in molecular weight is clearly evident upon ternary complex formation (Fig. 4, dashed arrows), further confirming the 2 : 1 : 2 solution stoichiometry.

Fig. 4 A log-log molecular weight calibration graph plotting D of commercially available HO-PEGs (M_n = 200–20,000 g mol⁻¹, blue) with an overlay of P1_free (M_n = 1100–8000 g mol⁻¹, green), and the 2 : 1 : 2 ternary complex upon CB[8] addition (M_n = 2200–16,000 g mol⁻¹, hollow red). D₂O at 27 °C (400 μM MVdimer in all samples).

Another observation during DOSY studies into P1 and P2 assembly with MVdimer-CB[8] (1 : 2) was the difference in the diffusion coefficient of MVdimer_free in the presence of P1 (log D_MVdimer = −9.53 m² s⁻¹, Fig. 2, top) compared with P2 (log D_MVdimer = −9.08 m² s⁻¹, Fig. 3), prior to any CB[8] addition. This was thought to be due to the different molecular weights of P1_5K and P2 (M_n = 19,000 g mol⁻¹) but further investigation found that MVdimer diffusion was independent of the surrounding polymer's molecular weight (Fig. S3, ESI†). The difference was in fact rationalised to be due to the inherently different environments felt by the TEG-based, tetracationic MVdimer in the presence of a PEG-based P1versus that of a PNIPAAM-based P2.

Solution viscosity studies in water were also performed, where a doubling in the molecular weight during ternary complex formation was expected to manifest itself as a reasonable increase in solution viscosity.^11,27,28 As shown in Fig. 5, the specific viscosities (η_sp) were measured for solutions containing MVdimer and either P1 (M_n = 1100–8000 g mol⁻¹, green circles) or HO-PEGs (M_n = 1100–12,000 g mol⁻¹, blue circles) in a 1 : 2 ratio. Ternary complex solutions upon CB[8] addition (M_n = 2200–16,000 g mol⁻¹, hollow red circles) were also plotted. A good overlap was observed between the viscosities of P1 and their unfunctionalised, HO-PEG analogues. Upon CB[8] addition, a doubling in molecular weight is indeed associated with a regular increase in viscosity as indicated by dashed arrows in Fig. 5. The ternary complexes (hollow red circles) exhibited slightly lower viscosities than predicted for their molecular weight along the PEG and P1 calibration plot. This reduction in viscosity was attributed to the effect of CB[8] addition; this was also observed upon adding increasing amounts of CB[8] to HO-PEGs of different molecular weights, where a marked reduction in reduced viscosity was observed (Fig. S8, ESI†).

Fig. 5 A log-log plot of specific viscosity (η_sp) against molecular weight for commercially available HO-PEGs (M_n = 1100–12,000 g mol⁻¹, blue) with an overlay of P1_free (M_n = 1100–8000 g mol⁻¹, green), and the 2 : 1 : 2 ternary complex upon CB[8] addition (M_n = 2200–16,000 g mol⁻¹, hollow red). H₂O at 30 °C (546 μM MVdimer in all samples).

In conclusion, we have shown the utility of DOSY NMR towards probing the multi-component assembly of supramolecular ABA triblock copolymers formed through CB[8] host–guest chemistry in aqueous solution. The adaptability of this method towards forming higher-ordered polymeric architectures stems from the versatility of the CB[8] host towards assembling a variety of guest-terminated polymers. Building up from the assembly of AB diblock copolymers⁷ and the characterisation of ABA triblock copolymer stability in the gas phase as demonstrated previously;⁸ we can now routinely use non-invasive, and in situ NMR methods to visualise the multi-component assembly process while deriving topological parameters such as molecular weight and solution mobility. Additionally, solution viscometry has provided further insight into the solution dynamics of the assembly components as they interact with each other upon ternary complex formation. Ongoing investigation will seek to develop the work reported here into further expanding CB[8]'s utility towards building stimuli-responsive, adaptable systems at the macromolecular interface, as well as probing the solution dynamics of the assemblies formed.

Acknowledgements

We gratefully acknowledge the financial support of the EPSRC and AWE. F. B. thanks the German Academic Exchange Service (DAAD) for financial support.

References

1. J. Lagona, S. Mukhopadhyay, Chakrabarti and L. Isaacs, Angew. Chem., Int. Ed., 2005, 44, 4844–4870.
2. J. W. Lee, S. Samal, N. Selvapalam, H.-J. Kim and K. Kim, Acc. Chem. Res., 2003, 36, 621–630.
3. H.-J. Kim, J. Heo, W. S. Jeon, E. Lee, J. Kim, S. Sakamoto, K. Yamaguchi and K. Kim, Angew. Chem., Int. Ed., 2001, 40, 1526–1529.
4. U. Rauwald, F. Biedermann, S. Deroo, C. V. Robinson and O. A. Scherman, J. Phys. Chem. B, 2010, 114, 8606–8615.
5. M. E. Bush, N. D. Bouley and A. R. Urbach, J. Am. Chem. Soc., 2005, 127, 14511–14517.
6. W. S. Jeon, E. Kim, Y. H. Ko, I. Hwang, J. W. Lee, S.-Y. Kim, H.-J. Kim and K. Kim, Angew. Chem., Int. Ed., 2005, 44, 87–91.
7. U. Rauwald and O. A. Scherman, Angew. Chem., Int. Ed., 2008, 47, 3950–3953.
8. S. Deroo, U. Rauwald, C. V. Robinson and O. A. Scherman, Chem. Commun., 2009, 644–646.
9. A. Harada, Y. Takashima and H. Yamaguchi, Chem. Soc. Rev., 2009, 38, 875–882.
10. H. Hofmeier, R. Hoogenboom, M. E. L. Wouters and U. S. Schubert, J. Am. Chem. Soc., 2005, 127, 2913–2921.
11. G. B. W. L. Lithgart, H. Ohkawa, R. P. Sijbesma and E. W. Meijer, J. Am. Chem. Soc., 2005, 127, 810–811.
12. J. D. Fox and S. J. Rowan, Macromolecules, 2009, 42, 6823–6835.
13. T. F. A. de Greef, M. M. J. Smulders, M. Wolffs, A. P. H. J. Schenning, R. P. Sijbesma and E. W. Meijer, Chem. Rev., 2009, 109, 5687–5754.
14. J. M. Zayed, N. Nouvel, U. Rauwald and O. A. Scherman, Chem. Soc. Rev., 2010, 39, 2806–2816.
15. A. Klaikherd, C. Nagamani and S. Thayumanavan, J. Am. Chem. Soc., 2009, 131, 4830–4838.
16. J. R. Capadona, K. Shanmuganathan, D. J. Tyler, S. J. Rowan and C. Weder, Science, 2008, 319, 1370–1374.
17. P. Cordier, F. Tournilhac, C. Soulie-Ziakovic and L. Leibler, Nature, 2008, 451, 977–980.
18. K.-S. Moon, H.-J. Kim, E. Lee and M. Lee, Angew. Chem., Int. Ed., 2007, 46, 1–5.
19. J.-H. Ryu, D.-J. Hong and M. Lee, Chem. Commun., 2008, 1043–1054.
20. C. Ott, R. Hoogenboom, S. Hoeppener, D. Wouters, J.-F. Gohy and U. S. Schubert, Soft Matter, 2009, 5, 84–91.
21. Y. Cohen, L. Avram and L. Frish, Angew. Chem., Int. Ed., 2005, 44, 520–554.
22. G. Groger, V. Stepanenko, F. Wurthner and C. Schmuck, Chem. Commun., 2009, 698–700.
23. T. Evan-Salem, I. Baruch, L. Avram, Y. Cohen, L. C. Palmer and J. Rebek, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 12296–12300.
24. W. S. Price, F. Elwinger, C. Vigouroux and P. Stilbs, Magn. Reson. Chem., 2002, 40, 391–395.
25. W. S. Price, Concepts Magn. Reson., 1998, 10, 197–237.
26. W. S. Price, Concepts Magn. Reson., 1997, 9, 299–336.
27. F. Huang, D. S. Nagvekar, X. Zhou and H. W. Gibson, Macromolecules, 2007, 40, 3561–3567.
28. A. Miyawaki, Y. Takashima, H. Yamaguchi and A. Harada, Tetrahedron, 2008, 64, 8355–8361.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details, synthesis of P1 and P2, NMR spectra of the different host–guest complexes (Figs S1–2), variation of D_MVdimer with PEG M_n (Fig. S3), ITC studies (Fig. S4–7, Table S1), CB[8]'s influence on PEG viscosity (Fig. S8). See DOI: 10.1039/c0py00197j

---