Driving force balance—the “identity card” of supramolecules in a self-sorting multicomponent assembly system

Abstract

Contrary to the popular belief that multicomponent assembly systems will theoretically co-assemble under the same type of driving forces, two distinct assembly modes from a system composed of two chemically similar supramolecules were demonstrated in this work. Although with exactly the same driving forces, molecule-level self-sorting unexpectedly occurred in this two-component system made of polyhedral oligomeric silsesquioxane (POSS) core-based supramolecules with one and eight lysine derivative arms. From the experiments, it was concluded that instead of driving force types, driving force counterpoise plays a vital role here, which we called “identity card hypothesis”. The hypothesis suggests that two highly similar components show high affinity for the same molecules through the differentiated “identity card”-like balance of driving forces induced by the difference in the molecular spatial shape, which has never been reported before.

---

Introduction

A self-sorting assembly, which with high-fidelity distinguishes “self” from “nonself”, is capable of self-assembling simultaneously and orthogonally in the multicomponent system.^1–9 It can be used to prepare multifunctional materials for various fields, such as drug-delivery systems,¹⁰ cell adhesives,¹¹ and photoelectric devices.¹² Depending on the process of system formation, self-sorting systems could be divided into thermodynamic self-sorting and kinetic self-sorting.¹³ Thermodynamic self-sorting occurs because of the structural differences between the components. When multiple components have diverse types of driving forces due to different structures, each component displays a high affinity for itself and the self-sorting assembly occurs. On the contrary, when the components have similar chemical structures, they are no longer able to distinguish themselves under the same kind of driving forces, then co-assembly is facilitated.¹⁴ D. K. Smith et al. have reported that the self-sorting assembly can be constructed by two dendritic molecules under different types of driving forces provided by diverse peripheral groups.¹⁵ However, self-sorting systems consisted of the components with similar structures that could be obtained by kinetic external stimuli such as pH changes¹⁶ and ultraviolet light,^17,18 while the disadvantage of kinetic self-sorting systems is that it is difficult to control accurately. Therefore, developing a simple method of self-sorting assembly in multicomponent systems with similar structures is crucial to extend the application of multifunctional self-sorting materials.

Supramolecular assembly systems can utilize a variety of non-covalent bond interactions to make molecules self-assemble into ordered and complex structures, which are ideal candidates for constructing self-sorting assemblies.^19–31 Among them, polyhedral oligomeric silsesquioxane (POSS) core-based supramolecules have attracted wide attention due to the unique three-dimensional (3D) cage-like POSS cores and a multi-branched outer “arm”, which endows them with highly controllable multiple peripheral groups and stereoscopic structures.^32–37 However, previous related studies have focused on peripheral groups of POSS core-based molecules, while the study of the contribution of POSS cores to the self-assembly is limited.^38–46 According to the studies reported in the literature, when the core is low-dimensional and non-stereometric, there is almost no stereo change in the molecular spatial shape, and the length of the arm and the number of peripheral groups have limited effects on the self-assembly of the molecules.^47,48 However, the POSS core is 3D cage-shaped with a strong stacking capacity; previous studies have not fully utilized the spatial structure and assembly capability of POSS cores for self-assembly studies. Thus, understanding and control of the influence of high-dimensional cores in a molecular shape and further in the self-assembly process is of great significance to complement the mechanism of the supramolecular assembly.

It is well known that the assembly mode of the same molecule can be tuned by affecting the competitive relationship between thermodynamic and kinetic processes.⁴⁹ The essence of this method is to affect the driving forces generated by thermodynamics or kinetics, so that the assembly mode is controlled by the dominant one. Similarly, for POSS core-based supramolecules with “cores” and peripheral “arms”, we hope that the ratio of cores to arms will be adjusted by modifying the molecular spatial shape. Furthermore, the proportion of the driving forces provided by cores and arms will change accordingly, and we will finally regulate the assembly mode of the molecule. Thus, PSS-Lys (shown in Fig. 1a and Scheme S1, ESI†) was designed and synthesized to obtain a typical tailored example to study the relationship between the molecular spatial shape and the assembly mode. Compared with the reported POSS core-based POSS-Lys (shown as Fig. 1b and Scheme S2, ESI†) with eight spatial symmetry lysine derivative arms,³⁸PSS-Lys has exactly the same POSS core and only one lysine arm, and the remaining seven outer arms are isobutyl groups. Compared with the peripheral groups of POSS-Lys, the isobutyl groups are small in volume and similar to lysine terminal groups, and the influence of steric hindrance and hydrophobicity on the assembly process can be neglected. Thus, the exposure degree of POSS cores is greatly increased, which could gain unneglectable solvophobic interactions by its 3D structure. Therefore, the modification of the molecular spatial shape caused significant alterations of the proportion of two types of driving forces provided by cores and arms. As a result, the entire driving forces counterpoise changes, and a new balance will be achieved, which is distinct from the one formed by POSS-Lys. By transforming the molecular spatial shape, 3D cage-like POSS cores can effectively amplify the difference in the counterpoise of driving forces between PSS-Lys and POSS-Lys, and two discrepant assembly modes were obtained. Under the same type of driving forces, the two-component assembly system composed of PSS-Lys and POSS-Lys was further studied, which should theoretically co-assemble. Surprisingly, each individual component maintained its assembly mode through the unique balance of driving forces, which acted like an “identity card” of molecules, and this system performed self-sorting without any external stimuli. Herein, a two-component assembly system specifically designed to study the effect of differences in driving forces balance on self-sorting is reported.

Fig. 1 Chemical structure of (a) PSS-Lys and (b) POSS-Lys.

Results and discussion

Assembly mode of PSS-Lys

In our previous research, the relationship between the driving forces and assembly modes was studied by performing coarse-grained molecular dynamics simulations⁴⁵ (Fig. S1, ESI†). It suggested that the strength of these two driving forces was affected by the variation in the solvent polarity, which further led to two diverse assembly behaviours. In this paper, the proportion of two driving forces was adjusted to affect the supramolecular assembly mode by modifying the molecular spatial shape. PSS-Lys was synthesized to be compared with POSS-Lys (Fig. S2–S6, ESI†), and the correlation of molecular spatial shape and assembly behaviours was obtained. In terms of chemical structures, the two supramolecules have the same driving forces for self-assembly: the POSS cores provide the solvophobic effect; the arms provide hydrogen bonding of carboxyl and amino groups. However, via modifying the molecular spatial shape of PSS-Lys, the ratio of cores and arms changed. Thus, the equilibrium between these two driving forces of PSS-Lys will be completely discrepant from that of POSS-Lys. Based on this assumption, the assembly mode of PSS-Lys at the distinct balance of driving forces was studied through the following experiments.

First, the gelation ability of PSS-Lys was examined through the tube inversion method in various organic solvents (Table 1). The results indicated that PSS-Lys was able to dissolve in various organic solvents of different polarities, resulting in a clear, transparent, non-viscous solution. In addition, it could not form a gel in these solvents even at a high concentration (80 mg mL⁻¹). By contrast, POSS-Lys only needed an extremely low concentration (0.5 mg mL⁻¹) to form a gel in the same organic solvents. It preliminarily indicated that the assembly behaviours of PSS-Lys were different from that of POSS-Lys. According to previous studies,^38–46 the hydrogen bonding of arms tended to self-assemble into fibers,^50,51 while the POSS cores tended to aggregate. Therefore, POSS-Lys with eight arms and one POSS core surrounded by arms had a higher proportion of hydrogen bonding than PSS-Lys, and the assembly behaviours tended to form a gel. While PSS-Lys had only a single lysine arm, the proportion of solvophobic effects generated by POSS cores increased, resulting in a decrease in the proportion of hydrogen bonds, so PSS-Lys failed to form a gel. It can be inferred that the driving force balance between the solvophobic effect and hydrogen bonding of PSS-Lys was inconsistent with POSS-Lys, resulting in the different assembly modes of PSS-Lys and POSS-Lys.

Table 1 Gelation ability of PSS-Lys in different solvents

Solvent	Polarity index	PSS-Lys
a The system is a solution. The concentration ranges from 0.5 to 80 mg mL^−1.
N-Hexane	0.06	S^a
Toluene	2.40	S
Methyl methacrylate	2.50	S
O-Dichlorobenzene	2.70	S
Styrene	3.00	S
Tetrahydrofuran	4.00	S
Ethyl acetate	4.40	S
Methanol	5.10	S

---

The assembly mode of PSS-Lys was further investigated by measuring the water contact angle of PSS-NH₂ (Fig. 2a) and PSS-Lys after grafting from a lysine derivative arm (Fig. 2b). The samples were prepared by dissolving the powders in the solvent, and the solvent was evaporated at room temperature. The water contact angle of PSS-NH₂ was 119.2°, and that of PSS-Lys was 68.3°. It turned out that PSS-NH₂ was hydrophobic, and PSS-Lys was hydrophilic. As the POSS inorganic core consists of Si–O, an alternately connected silicon oxide skeleton was highly hydrophobic, and PSS-NH₂ tended to be hydrophobic when the arms were hydrophilic amino groups, which indicated that the amino groups were distributed inside the assembly, while the POSS cores were exposed outside. This is due to the short length of the amino groups, which have little steric effects on POSS cores. POSS cores were exposed outside to a great extent, and its strong solvophobic effect led to agglomeration, forming an assembly with the cores outside and the arms inside. For PSS-Lys, when the highly hydrophobic POSS cores were clustered inside the assembly and the hydrophilic lysine arms were at the periphery, the assemblies were generally hydrophilic. This is because the lysine arms had a stronger steric effect than amino groups, which weakened the effect of the POSS cores. In addition, the lysine arms also provided hydrogen bonding, which formed a competitive relationship with the driving forces of POSS cores. The results were the assemblies with the cores inside and arms outside. Therefore, it can be speculated that PSS-Lys will tend to form the aggregates with POSS cores inside and arms outside.

Fig. 2 Contact angle of the film formed by (a) PSS-NH₂ and (b) PSS-Lys after grafting from Lys.

The assembly mode of PSS-Lys was compared with that of POSS-Lys. In addition, the differences between the two assembly morphologies were captured by scanning electron microscopy (SEM). In the methyl methacrylate (MMA), the morphology of POSS-Lys was a unique loofah-like fiber network (Fig. S7, ESI†). However, the morphology of PSS-Lys was irregularly lumpy aggregates (Fig. 3a), indicating that PSS-Lys tended to aggregate rather than assemble into fibers, unlike POSS-Lys, which also verified the previous experimental results. In terms of chemical structures, the hydrogen bonding provided by arms is very important, but due to the change in the molecular spatial shape, the solvophobic effect of POSS cores is also critical. POSS-Lys has eight lysine derivative arms, which result in stronger hydrogen bonding. Moreover, because of the steric hindrance effect, the POSS cores cannot approach each other, and hence, the eight lysine arms provided a fixation effect. POSS-Lys riveted together in a “face-to-face” stack to form a vertical assembly, and finally, the loofah-like network was obtained (Fig. S8, ESI†). However, PSS-Lys only has a single arm, and the hydrogen bonding is greatly weakened compared to POSS-Lys. Meanwhile, the steric hindrance effect of a single arm is so weak that the POSS cores cannot be fixed, and the solvophobic effect of POSS cores is amplified. Thus, in the PSS-Lys system, the balance of the two driving forces had changed significantly when compared to POSS-Lys, resulting in two completely different assembly modes. The XRD results showed that the 2θ value corresponding to the characteristic peaks of PSS-Lys was larger than that of POSS-Lys (Fig. 3b). According to Bragg's law, when the value of 2θ became larger, the crystal plane distance became smaller, which indicated that the distance between the POSS cores in the PSS-Lys assembly was closer than that in POSS-Lys, and PSS-Lys tended to agglomerate. In addition, because the strong solvophobic effect of POSS cores towards crystallization was dominant, compared to POSS-Lys, the characteristic peaks of PSS-Lys were much sharper, manifesting that the assembly of PSS-Lys was more orderly, which further validated the assembly of PSS-Lys. In conclusion, by changing the molecular spatial shape, the 3D POSS cores amplified the difference in the balance of driving forces between PSS-Lys and POSS-Lys, and two distinct assembly modes of the two molecules with similar chemical structures were obtained.

Fig. 3 (a) TEM images of PSS-Lys. (b) The XRD figure of PSS-Lys and POSS-Lys.

Assembly mode of the system composed of PSS-Lys and POSS-Lys

The next level of complexity involved investigating the assembly behaviours of the two-component assembly system composed of PSS-Lys and POSS-Lys. The concentration of POSS-Lys in this system was set to its minimal gel concentration (0.5 mg mL⁻¹). On the premise that POSS-Lys was assembled into the fiber network, PSS-Lys was introduced into the system at different molar ratios (PSS-Lys : POSS-Lys = 1 : 1, 4 : 1, and 8 : 1).

In order to study the assembly mode of this two-component assembly system, the gelation ability of this system in different proportions was tested by the tube inversion method in major gelation solvents (Table 2). The mixture systems were still capable of forming a transparent gel in gelation solvents, the gelation ability of POSS-Lys did not change significantly with the addition of PSS-Lys. In addition, the gel–sol transition temperature (T_gel) of the POSS-Lys gel and PSS-Lys/POSS-Lys mixed gel was assessed at different molar ratios (Fig. S9 and S10, ESI†). It can be seen that the T_gel values of the POSS-Lys and PSS-Lys/POSS-Lys gels at different concentrations were almost the same (40 ± 1 °C, 58 ± 1 °C, 70 ± 1 °C, 76 ± 1 °C, and 78 ± 1 °C). Clearly, PSS-Lys, which could not form a gel, did not appear to disrupt the gelation ability of POSS-Lys. This phenomenon is consistent with the fact that the gelator with a weaker gelation ability in the self-sorting assembly cannot affect another gelator with strong gelation ability.¹⁵ Furthermore, microscopic observation is used to verify the assembly mode of this system, which is the most intuitionistic and effective way for assembly research. Interestingly, regardless of the PSS-Lys : POSS-Lys ratio, the morphology always turned out to be a mixture of two microstructures (irregular lumpy aggregates and loofah-like fibers) without the distinguishable difference in the SEM and TEM images (Fig. 4a–f). Each component maintained their individual assembly mode in this system. Moreover, as the concentration of PSS-Lys increased, the number and size of aggregates of PSS-Lys were getting larger, from 0.7 μm (PSS-Lys : POSS-Lys = 1 : 1, mol : mol) to 1.4 μm (PSS-Lys : POSS-Lys = 4 : 1, mol : mol), and then to 1.8 μm (PSS-Lys : POSS-Lys = 8 : 1, mol : mol). The surplus PSS-Lys gel still shows high affinity for itself instead of co-assembling with POSS-Lys. Therefore, the microscopic images shown in Fig. 4 can be considered a self-sorting assembly. Moreover, by analysing the energy-dispersive X-ray spectra of white lumpy aggregates and loofah-like fiber networks in this mixture system consisting of PSS-Lys : POSS-Lys = 1 : 1 (mol : mol) shown in Fig. 4a, the results indicated that PSS-Lys contains 51.6% C,4.9% N, 24.0% O, and 19.5% Si and POSS-Lys contains 57.0% C, 10.7% N, 25.2% O, and 7.1% Si (Fig. S11, ESI†). The element content of white lumpy aggregates was basically consistent with the theoretical content of PSS-Lys (C: 47.3%; N: 4.5%; O: 22.0%; Si: 18.0%; and H: 8.4%), and the element content of loofah-like fiber networks was consistent with the theoretical content of POSS-Lys (C: 52.0%; N: 9.6%; O: 23.7%; Si: 6.4%; and H: 8.3%). It is obvious that there was a significant difference in the content of nitrogen and silicon, which proved that the aggregation of PSS-Lys dispersed individually in the loofah-like fiber networks of POSS-Lys. These phenomena further proved intuitively that self-sorting occurred under the same type of driving forces in the two-component assembly system.

Table 2 Gelation ability of the PSS-Lys/POSS-Lys two-component system in different solvents

Gelation solvent	1 : 1	4 : 1	8 : 1	POSS-Lys	PSS-Lys
a The system is a transparent gel. b The system is a solution.
Toluene	TG^a	TG	TG	TG	S^b
Methyl methacrylate	TG	TG	TG	TG	S
O-Dichlorobenzene	TG	TG	TG	TG	S
Styrene	TG	TG	TG	TG	S
Ethyl acetate	TG	TG	TG	TG	S

---

Fig. 4 SEM images and TEM images of the PSS-Lys/POSS-Lys two-component system. (a, c and e) SEM images of the PSS-Lys : POSS-Lys = 1 : 1, 4 : 1, and 8 : 1 (mol : mol) xerogel, 5 mg mL⁻¹. (b, d and f) TEM images of the PSS-Lys : POSS-Lys = 1 : 1, 4 : 1, and 8 : 1 (mol : mol) xerogel, 5 mg mL⁻¹.

The assembly mode of this system was also studied at the molecular level by Fourier-transform infrared (FT-IR) spectroscopy.⁵² The FT-IR spectra can provide insights into the molecular interactions such as hydrogen bonding between carboxyl and amino groups in this complex system, which is one of the major driving forces in this system. It has been reported that, if a system were self-sorted, then we would expect no change in the absorption of the individual components and the combined spectrum should be similar to the overlay of the two spectra for the individual components.⁵³ However, when the co-assembly occurs, this could be seen directly in the FT-IR spectra, with the appearance of new peaks and the shift of original peaks.⁵⁴ Herein, the IR spectra of PSS-Lys, POSS-Lys xerogel and xerogel of PSS-Lys : POSS-Lys = 1 : 1, 4 : 1, and 8 : 1 (mol : mol) were obtained (Fig. 5). Since the hydrogen bonding effect is one of the major driving forces of this system, we focused on the peak shift for carboxyl and amino groups, which can form hydrogen bonding, and the appearance of the new peaks to probe the packing mode of the two-component gel system. As shown in Fig. 5a, there was almost no shifting in the stretching vibration peaks of N–H from amino groups (ν_N–H = 3349 cm⁻¹) and C=O from carboxyl groups (ν_C=O = 1710 cm⁻¹). If the co-assembly occurs, there will be a hydrogen bond interaction between the two components, then the infrared absorption peaks of N–H and C=O will inevitably shift to a lower wavenumber. In addition, there was no new peak in this spectra, for example, in Fig. 5b, both the single stretching vibration peak of C–H of PSS-Lys (ν_C–H = 2955 cm⁻¹) and the double peaks of POSS-Lys (ν_C–H = 2977 cm⁻¹ and ν_C–H = 2932 cm⁻¹) still existed in the two-component gel system without any shifting, regardless of the ratio of PSS-Lys and POSS-Lys. In the two-component system, the stretching vibration peaks of C–H were only the superposition of the peaks of PSS-Lys and POSS-Lys. Therefore, the spectra of the mixed system were simply an overlay of the spectra for the two components. Based on the results of FT-IR, the assembly mode of this two-component system was self-sorting.

Fig. 5 (a) FT-IR spectra of PSS-Lys, POSS-Lys and PSS-Lys : POSS-Lys = 1 : 1, 4 : 1, and 8 : 1 (mol : mol) xerogel obtained from MMA. (b) Amplified image of the dashed box in (a).

The rheological measurements are also crucial to understand the self-sorting assembly system, when the concentration of PSS-Lys remained unchanged (Fig. 6). It was found that with the increase in the concentration of POSS-Lys, the storage modulus (G′) and loss modulus (G′′) gradually increased, which meant that POSS-Lys self-assembled into a denser and stabler nanofiber network with the increase in concentration. This proved that the assemblies of POSS-Lys were dominant in this two-component system.

Fig. 6 Frequency dependence of G′ and G′′ for PSS-Lys/POSS-Lys system, measured at 25 °C with a frequency from 0.1 to 100 rad s⁻¹ and strain g = 0.1%. PSS-Lys/POSS-Lys with 1.4% : 0.5% and 1.4% were prepared in MMA.

Similarly, when the concentration of POSS-Lys was constant (Fig. 7a and b). The data showed that the mechanical strength of this system was almost unchanged with the increase in concentration of PSS-Lys. As shown in SEM images (Fig. 4), the size of PSS-Lys assemblies was much smaller than that of POSS-Lys. In addition, the assembly mode of this system was self-sorting, which meant that the two components self-assembled separately. Therefore, the increase in the PSS-Lys aggregates had almost no effect on the mechanical strength of the entire system. Hence, the rheological results proved that the self-sorting assembly occurred under the same kind of driving forces.

Fig. 7 Frequency dependence of G′ and G′′ for the PSS-Lys/POSS-Lys system, measured at 25 °C with a frequency from 0.1 to 100 rad s⁻¹ and strain g = 0.1%. (a) PSS-Lys/POSS-Lys with 2.8% : 1.0% and 1.4% were prepared in MMA. (b) PSS-Lys/POSS-Lys with 1.4% : 0.5% and 0.7% were prepared in MMA.

Based on this anomalous phenomenon of self-sorting in this assembly system composed of two similar chemical structural components, PSS-Lys and POSS-Lys, one hypothesis was proposed: under multiple driving forces, each supramolecule will have a unique balance of driving forces in the assembly process. This balance of driving forces can be considered as a molecular “identity card”. In this system, the “identity card” was the balance of the solvophobic effect provided by POSS cores and the hydrogen bonding provided by lysine arms, the counterpoise of which plays a key role instead of each one individually. As shown in Fig. 8, due to the difference in the molecular spatial shape, the proportions of driving forces in the two molecules changed greatly, which made the equilibrium of driving forces unique to each component. The distinction of driving force balance between PSS-Lys and POSS-Lys led to their very differentiated assembly mode. It is this “identity card” that enabled them to show high affinity for the same molecules without any interference with other molecules and eventually forming the self-sorting assembly in the complex two-component system.

Fig. 8 Schematic of the PSS-Lys/POSS-Lys two-component assembly system.

Conclusions

In this paper, the assembly behaviours of PSS-Lys with one lysine derivative arm and POSS-Lys with eight arms were studied. By changing the molecular spatial shape, the 3D cage-like POSS cores magnified the difference in the balance of driving forces provided by the POSS cores and lysine arms. Through a series of tests, it was proved that two distinct assembly modes of PSS-Lys and POSS-Lys with similar chemical structures can be obtained. Furthermore, a two-component assembly system was constructed from these two supramolecules with similar chemical structures to study the assembly mode of this system under the same type of driving forces. Unexpectedly, a series of studies had shown that this two-component system self-sorted, which theoretically tended to form a co-assembly. Moreover, two highly similar components showed high affinity for the same molecules through the differentiated “identity card”-like equilibrium relationship between two kinds of driving forces, and hence, the precise self-sorting occurred in the complicated two-component system. It provides a new strategy for the precise self-sorting mechanism without any external stimuli, which helps develop multifunctional self-sorting supramolecular materials.

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. J. M. Lehn, Science, 1985, 227, 849–856.
2. J. M. Lehn, Science, 1993, 260, 1762–1763.
3. M. C. LeMieux, M. Roberts, S. Barman, Y. W. Jin, J. M. Kim and Z. Bao, Science, 2008, 321, 101–104.
4. M. M. Safont-Sempere, G. Fernandez and F. Wurthner, Chem. Rev., 2011, 111, 5784–5814.
5. L. E. Buerkle and S. J. Rowan, Chem. Soc. Rev., 2012, 41, 6089–6102.
6. S. Liu, C. Ruspic, P. Mukhopadhyay, S. Chakrabarti, P. Y. Zavalij and L. Isaacs, J. Am. Chem. Soc., 2005, 127, 15959–15967.
7. F. Wang, C. Y. Han, C. L. He, Q. Z. Zhou, J. Q. Zhang, C. Wang, N. Li and F. H. Huang, J. Am. Chem. Soc., 2008, 130, 11254–11255.
8. J. F. Ayme, J. M. Lehn, C. Bailly and L. Karmazin, J. Am. Chem. Soc., 2020, 142, 5819–5824.
9. J. Deng, D. Bezold, H. Jessen and A. Walther, Angew. Chem., Int. Ed., 2020, 132, 12182–12190.
10. Y. Yang, H. Hu, L. Chen, H. Bai, S. Wang, J. F. Xu and X. Zhang, Mater. Chem. Front., 2019, 3, 806–811.
11. M. Zhou, A. M. Smith, A. K. Das, N. W. Hodson, R. F. Collins, R. V. Ulijn and J. E. Gough, Biomaterials, 2009, 30, 2523–2530.
12. K. Sugiyasu, S. I. Kawano, N. Fujita and S. Shinkai, Chem. Mater., 2008, 20, 2863–2865.
13. A. X. Wu and L. Isaacs, J. Am. Chem. Soc., 2003, 125, 4831–4835.
14. J. Raeburn and D. J. Adams, Chem. Commun., 2015, 51, 5170–5180.
15. J. R. Moffat and D. K. Smith, Chem. Commun., 2009, 316–318.
16. S. Bloom, C. Liu, D. K. Kolmel, J. X. Qiao, Y. Zhang, A. M. Poss, W. R. Ewing and D. W. C. MacMillan, Nat. Chem., 2018, 10, 205–211.
17. D. J. Cornwell, O. J. Daubney and D. K. Smith, J. Am. Chem. Soc., 2015, 137, 15486–15492.
18. K. L. Morris, L. Chen, J. Raeburn, O. R. Sellick, P. Cotanda, A. Paul, P. C. Griffiths, S. M. King, R. K. Serpell and D. J. Adams, Nat. Commun., 2013, 4, 1480.
19. J. M. Lehn, Science, 2002, 295, 2400–2403.
20. J. M. Lehn, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 4763–4768.
21. J. M. Lehn, Chem. Soc. Rev., 2017, 46, 2378–2379.
22. A. Sarkar, R. Sasmal, C. Empereur-mot, D. Bochicchio, S. V. K. Kompella, K. Sharma, S. Dhiman, B. Sundaram, S. S. Agasti, G. M. Pavan and S. J. George, J. Am. Chem. Soc., 2020, 142, 7606–7617.
23. L. He, S. C. Wang, L. T. Lin, J. Y. Cai, L. Li, T. H. Tu and Y. T. Chan, J. Am. Chem. Soc., 2020, 142, 7134–7144.
24. Q. Shi, X. Zhou, W. Yuan, X. Su, A. Neniskis, X. Wei, L. Taujenis, G. Snarskis, J. S. Ward, K. Rissanen, J. de Mendoza and E. Orentas, J. Am. Chem. Soc., 2020, 142, 3658–3670.
25. H. Barbero, N. A. Thompson and E. Masson, J. Am. Chem. Soc., 2020, 142, 867–873.
26. K. Aratsu, R. Takeya, B. R. Pauw, M. J. Hollamby, Y. Kitamoto, N. Sihmizu, H. Takagi, R. Haruki, S. I. Adachi and S. Yagai, Nat. Commun., 2020, 11, 1623.
27. Y. Li, C. Liu, X. Bai, F. Tian, G. Hu and J. Sun, Angew. Chem., Int. Ed., 2020, 59, 3486–3490.
28. Y. Wang, T. K. Piskorz, M. Lovrak, E. Mendes, X. Guo, R. Eelkema and J. H. van Esch, Adv. Sci., 2020, 7, 1902487.
29. X. Shi, X. Zhang, X. L. Ni, H. Zhang, P. Wei, J. Liu, H. Xing, H. Q. Peng, J. W. Y. Lam, P. Zhang, Z. Wang, H. Hao and B. Z. Tang, Macromolecules, 2019, 52, 8814–8825.
30. L. Volbach, N. Struch, F. Bohle, F. Topic, G. Schnakenburg, A. Schneider, K. Rissanen, S. Grimme and A. Lutzen, Chemistry, 2020, 3335–3347.
31. Y. T. Tsai, G. Raffy, H. F. Liu, B. J. Peng, K. P. Tseng, L. Hirsch, A. Del Guerzo, D. M. Bassani and K. T. Wong, Mater. Chem. Front., 2020, 4, 845–850.
32. M. A. Castriciano, N. Leone, P. Cardiano, S. Manickam, L. M. Scolaro and S. Lo Schiavo, J. Mater. Chem. C, 2013, 1, 4746–4753.
33. M. B. Hu, Z. Y. Hou, W. Q. Hao, Y. Xiao, W. Yu, C. Ma, L. J. Ren, P. Zheng and W. Wang, Langmuir, 2013, 29, 5714–5722.
34. H. K. Shin, C. C. Hsieh, M. G. Mohamed, C. Y. Zhu and S. W. Kuo, Soft Matter, 2016, 12, 1847–1858.
35. G. Tang, X. Wang, D. Li, Y. Ma and D. Wu, Polym. Chem., 2018, 9, 5377–5384.
36. L. Fan, X. Wang, Q. Cao, Y. Yang and D. Wu, Biomater. Sci., 2019, 7, 1984–1994.
37. N. Lu, Y. Lu, S. Liu, C. Jin, S. Fang, X. Zhou and Z. Li, Biomacromolecules, 2019, 20, 3485–3493.
38. G. Tang, S. Chen, F. Ye, X. Xu, J. Fang and X. Wang, Chem. Commun., 2014, 50, 7180–7183.
39. S. Chen, X. Luo, H. He, X. Tong, B. Wu, M. Ma and X. Wang, J. Mater. Chem. C, 2015, 3, 12026–12031.
40. H. He, S. Chen, J. Bai, H. Zheng, B. Wu, M. Ma, Y. Shi and X. Wang, RSC Adv., 2016, 6, 34685–34691.
41. H. He, S. Chen, X. Tong, Y. Chen, B. Wu, M. Ma, X. Wang and X. Wang, Soft Matter, 2016, 12, 957–964.
42. S. Chen, X. Tong, H. He, M. Ma, Y. Shi and X. Wang, ACS Appl. Mater. Interfaces, 2017, 9, 11924–11932.
43. Z. An, S. Chen, X. Tong, H. He, J. Han, M. Ma, Y. Shi and X. Wang, Langmuir, 2019, 35, 2649–2654.
44. X. Tong, T. Shan, Q. Du, Z. An, H. He, M. Ma, Y. Shi, S. Chen and X. Wang, Adv. Electron. Mater., 2019, 5, 1900373.
45. T. Shan, L. Gao, X. Tong, Q. Du, Z. An, H. He, J. Lin, S. Chen and X. Wang, Chem. – Eur. J., 2019, 25, 12098–12104.
46. H. He, H. Zheng, M. Ma, Y. Shi, Z. Gao, S. Chen and X. Wang, Chem. Commun., 2020, 56, 2983–2986.
47. V. Percec, P. W. Chu, G. Ungar and J. P. Zhou, J. Am. Chem. Soc., 1995, 117, 11441–11454.
48. R. Zhang, X. Feng, R. Zhang, W. Shan, Z. Su, J. Mao, C. Wesdemiotis, J. Huang, X. Y. Yan, T. Liu, T. Li, M. Huang, Z. Lin, A. C. Shi and S. Z. D. Cheng, Angew. Chem., Int. Ed., 2019, 58, 11879–11885.
49. J. Matern, Y. Dorca, L. Sanchez and G. Fernandez, Angew. Chem., Int. Ed., 2019, 58, 16730–16740.
50. D. Yan, G. R. Williams, M. Zhao, C. Li, G. Fan and H. Yang, Langmuir, 2013, 29, 15673–15681.
51. J. Puigmarti-Luis, A. Minoia, A. P. del Pino, G. Ujaque, C. Rovira, A. Lledos, R. Lazzaroni and D. B. Amabilino, Chem. – Eur. J., 2006, 12, 9161–9175.
52. Y. M. Abul-Haija, S. Roy, P. Frederix, N. Javid, V. Jayawarna and R. V. Ulijn, Small, 2014, 10, 973–979.
53. K. Sugiyasu, S. I. Kawano, N. Fujita and S. Shinkai, Chem. Mater., 2008, 20, 2863–2865.
54. P. Xing, X. Chu, S. Li, F. Xin, M. Ma and A. Hao, New J. Chem., 2013, 37, 3949–3955.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0sm01405b

‡ Si Chen and Likang Zhou contributed equally.

---