Two kinds of guest molecules, methoxy-tetraphenylethylene derivatives with one (TPE-PY) or two (TPE-DPY) flexible alkyl-bridged phenyl pyridines groups were synthesized by Mizoroki-Heck reaction and alkyl substitution reaction, which were characterized through nuclear magnetic resonance (1H NMR, 13C NMR) and high-resolution mass spectrometry (HRMS) (Supplementary Figs. 2–10). A series of reference molecules, including alkyl-chain-modified bromophenylpyridinium salts (PY-1), tetraphenylethene derivative possessing vinyl pyridine salts (TPE-1, TPE-2) (Supplementary Fig. 1 and Supplementary Figs. 11–13) were synthesized to process the relevant control experiments for exploring the binding mode of guest molecules with CB[n] (n = 7/8) and the single-molecule PRET luminescence behavior. In contrast to the previously reported mono-bromophenylpyridine derivatives,39,40 the guest molecule TPE-DPY has two alkyl-bridged bipyridine salt units, which provide more binding sites to assemble with CB[n] (n = 7/8) through ionic dipole interaction and hydrophobic interaction. First, 1H NMR experiments (Supplementary Fig. 14) and 2D correlation spectroscopy (COSY) (Supplementary Fig. 15) were conducted to investigate the binding behavior between TPE-DPY and CB[8]. In Supplementary Fig. 14, upon the addition of CB[8] into the guest solution, the proton signal of TPE-DPY gradually passivated and no longer changed when the CB[8] concentration exceeded 1 equivalent, indicating the complexation of TPE-DPY and CB[8] reached an equilibrium stage. Similarly, the UV titration spectrum of TPE-DPY exhibited a persistent red-shift until stabilizing around 1.0 equivalent CB[8], and the related binding constant was obtained as 5.50 × 106 M− 1 (Supplementary Fig. 16). Job’s plot measured by UV-vis spectra confirmed a 1:1 stoichiometric ratio of TPE-DPY to CB[8] (Supplementary Fig. 17). Furthermore, two-dimensional rotating frame overhauser effect spectroscopy (ROESY) (Supplementary Fig. 18) and two-dimensional diffusion-ordered spectroscopy (DOSY) (Supplementary Fig. 19) were carried out to infer the binding mode. The correlation signals of proton Hb’ and Hg’ in TPE-DPY (Supplementary Fig. 18) manifested a deep encapsulation of bromophenylpyridine units by CB[8] cavity in a head-to-tail binding mode. The diffusion coefficients of guest molecule TPE-DPY (D = 2.09×10− 10 m/s2) and assembly TPE-DPY/CB[8] (D = 5.19×10− 11 m/s2) differed by an order of magnitude (Supplementary Fig. 19), which verified the formation of n:n head-to-tail chain supramolecular pseudorotaxane for TPE-DPY/CB[8].41 Correspondingly, transmission electron microscopy (TEM) and scanning electron microscopy (SEM) experiments revealed that the free guest molecule TPE-DPY presented ellipsoidal-shaped nanoparticles with sizes ranging from 50–90 nm (Figs. 3a, 3e, 3i), owing to the hydrophobic interaction and the stacking of tetraphenylethlene groups. In comparison, the TPE-DPY/CB[8] complex formed a nanorod with a length of approximately 500 nm (Figs. 3b, 3f, 3j), consistent with the head-to-tail chain pseudorotaxane assembly mode.
Unlike the complexation of TPE-DPY and CB[8], Job’s plot determined by UV-vis spectra of TPE-DPY and CB[7] showed the inflection point at 0.2, implying a 1:4 stoichiometry ratio for TPE-DPY/CB[7] (Supplementary Fig. 20a). Nevertheless, no meaningful information was captured in 1H NMR titration experiments of TPE-DPY and CB[7] because of the strong proton peak passivation following the addition of CB[7] (Supplementary Fig. 21). Therefore, TPE-PY, consisting of the same functional groups as TPE-DPY, was selected as a reference compound for controlled experiments. Combined the 1H NMR titration spectrum (Fig. 2a and Supplementary Fig. 22) and 2D COSY (Supplementary Figs. 23–24), we found that upon the increasing amount of CB[7] from 0–1 equivalent, an apparent high-field shift of Ha and Hb on the bromophenylpyridine group was observed resulting from the shielding effect, while the protons Hc, Hd and He shifted download indicating that ethylene pyridine moiety was located outside of CB[7]. As the addition of CB[7] excessed 1 equivalent, Ha and Hb underwent a slight shift towards the low field concurrently accompanied by an up-field shift of Hc and Hd. It indicates that CB[7] preferentially binds to the bromophenylpyridine section at a low concentration and then assembles with the ethylene pyridine unit at a high concentration. The binding constants obtained through UV titration experiments give more proof for the above results, in which the binding constants of PY-1/CB[7] were brought to be 9.17 × 106 M− 1 higher than 2.72 × 105 M− 1 of TPE-2/CB[7] (Fig. 2b). Moreover, Job’s plot demonstrated a 1:2 stoichiometric ratio of TPE-PY to CB[7] (Supplementary Fig. 20b), further confirming the 1:4 binding mode between TPE-DPY and CB[7]. The assembly of TPE-DPY/CB[7] was visible in the TEM and SEM images as nanospheres with the increased particle size caused by the rising CB[7] concentration. As shown in Fig. 3 and Supplementary Fig. 29, a nanosphere measuring roughly 100 nm in diameter was produced by adding 2 equivalents of CB[7] and a large size nanosphere of about 200 nm developed as the amount of CB[7] increased to 4 equivalents, owing to the binding effect of CB[7] for ethylene pyridine portions and bromophenylpyridine groups, which increased the rigidity of supramolecular assembly and the space for stacking arrangement (Figs. 3c, 3g, 3k).
Interestingly, entirely varying from the binary assembly of TPE-DPY/CB[7] and TPE-DPY/CB[8], the morphology of TPE-DPY/CB[7]/CB[8] co-assembly exhibited a polyhedral nanoplate with distinct edges and corners in TEM images (Figs. 3d and 3h). Under a scanning electron microscope, it was evident that the three-dimensional nanoplates were formed by hierarchical self-assembly (Fig. 3l and Supplementary Fig. 29c), which resulted from the side-by-side and layer-by-layer stacking of TPE-DPY/CB[7]/CB[8] supramolecular assemblies with a sizeable rigid core and soft chains. The analysis of the co-assembly of the reference molecule TPE-PY with CB[7] and CB[8] allowed us to infer the binding mode of the ternary assembly TPE-DPY/CB[7]/CB[8]. Specifically, 1H NMR titration experiment showed that on the basis of TPE-PY/CB[7] with a stoichiometric ratio of 1:1 where CB[7] bound to the phenylpyridine unit, vinylpyridine protons (Hc, Hd, and He) exerted an up-field shift upon increase CB[8] concentration from 0 to 0.5 equivalent, indicating the tight encapsulation of ethylene pyridine moiety by CB[7] that moved from the phenylpyridine unit (Fig. 2a and Supplementary Figs. 25–26). Moreover, in the 2D NOESY spectrum of the TPE-PY/CB[7]/CB[8] assembly (Supplementary Fig. 27), we could easily locate the cross-peaks between the protons of vinyl functional groups and CB[7]. The 2:1 stoichiometry ratio obtained from Job’s plot (Supplementary Fig. 28) and the strong binding constant of 2.90×1012 M− 2 for TPE-PY/CB[8] (Fig. 2b) provided further support for the aforementioned findings. These findings suggested that the vinylpyridine and phenyl pyridines moieties of TPE-PY were included in the cavities of CB[7] and CB[8], respectively, ultimately generating a ternary supramolecular assembly with a stoichiometric ratio of TPE- PY:CB[7]:CB[8] = 2:2:1. Thus, we deduced that the co-assembly of TPE-DPY/CB[7]/CB[8] went through a similar assembly process to form a linear supramolecular aggregate with a stoichiometric ratio of 1:2:1. From the above experimental results, it can be seen that CB[n] (n = 7, 8) possesses a different binding affinity to TPE-DPY, leading to a diverse topological morphology for the supramolecular assembly. Profited by the host-guest complexation, hydrophobic interaction, and π – π stacking interactions, the binary assembly of TPE-DPY/CB[7] presents spherical nanoparticles with adjustable dimensions, and CB[8] with a larger hydrophobic cavity binds with TPE-DPY to form a n:n rod-shaped pseudorotaxane. The ternary co-assembly TPE-DPY/CB[7]/CB[8] has a more robust rigid structure than TPE-DPY/CB[7] and TPE-DPY/CB[8], resulting in a linear self-assembly stacked multi-layered three-dimensional nanoplates.
Subsequently, the configuration-confined photophysical properties of the assembly of TPE-DPY and CB[n] (n = 7/8) were explored. With the binary assembly of CB[7] or CB[8], the absorption peak of TPE-DPY redshifted from 315 nm to 320 nm and 330 nm, respectively, and a comparable red shift by 13 nm occurred in the co-assembly of CB[7] and CB[8] (Fig. 2c). For the photoluminescence spectra shown in Figs. 3a and 3b, the guest molecule TPE-DPY exhibited a fluorescence emission at 390 nm with an excitation of 315 nm, and no phosphorescent signal was captured in the delayed spectrum. The assembly of TPE-DPY/2CB[7], TPE-DPY/4CB[7], TPE-DPY/CB[8], and TPE-DPY/CB[7]/CB[8] displayed a remarkable new emission peak around 530 nm as compared to the free TPE-DPY (Fig. 4a). Differing from the steady-state PL spectrum spectra, the delay spectra of these assemblies showed a major emission peak near 530 nm, illustrating its long-lived feature, which was further proved by their microsecond lifetime obtained by the time decay curve measurement (Figs. 4b, c). Notably, in contrast to the binary assembly TPE-DPY/2CB[7], TPE-DPY/4CB[7], and TPE-DPY/CB[8], the ternary assembly TPE-DPY/CB[7]/CB[8] had a stronger luminous intensity with extending the lifetime from 8.29 µs to 59.36 µs (Fig. 4c), because of the synergistic confinement effect of CB[7] and CB[8] on guest molecules and the supramolecular nanostructure formed by the linear rigid assembly layer by layer enabling valid shielding effect on the quencher. Additionally, after the injection of Ar, the lifetime of TPE-DPY/CB[7]/CB[8] aqueous solution at 540 nm was significantly increased from 59.36 µs to 129.97 µs (Supplementary Fig. 30) due to the avoidance of the triplet electron quenching caused by oxygen, further confirming the phosphorescence properties of emission peak at 540 nm. The above experiment results demonstrated that the macrocyclic confinement can effectively induce a phosphorescence emission, and the topological morphology of supramolecule assembly can be regulated by adjusting the ratios of CB[7] and CB[8], presenting different photophysical properties.
On the basis of the binary supramolecular assembly, the multivalent cascade assembly has evolved into an effective method to improve phosphorescence performance.42 Herein, HACD as a polysaccharide targeting agent has been introduced into TPE-DPY/CB[7]/CB[8] to construct a secondary assembly, which resulted in an exchanged topology from hierarchical self-assembled nanoplates to spherical nanoparticles (Figs. 3d and 5a). Dynamic light scattering (DLS), TEM, and zeta potential experiments were carried out to explore the assembly behavior for TPE-DPY/CB[7]/CB[8]@HACD. DLS measurements suggested that TPE-DPY/CB[7]/CB[8]@HACD assembly had an average hydrodynamic diameter of 236 nm, which matched the size of nanospheres in the TEM image (Figs. 5a, b). Moreover, on the contrary of TPE-DPY and TPE-DPY/CB[7]/CB[8] that possessed a positive zeta potential at + 1.74 and + 1.80 mV, respectively, a negative potential value of TPE-DPY/CB[7]/CB[8]@HACD was obtained as -0.335 mV (Supplementary Fig. 31), revealing the successful construction of the multicomponent assembly. Remarkably, the cascade assembly of TPE-DPY/CB[7]/CB[8]@HACD not only changed the topological morphology but also achieved a HACD-activated single-molecular PRET based on macrocyclic confinement. As shown in Figs. 5e, 5f and Supplementary Fig. 32, upon the addition of HACD, the assembly of TPE-DPY/CB[7]/CB[8] excited by 333 nm showed a weak phosphorescence at 535 nm with a lifetime of 69.83 µs, and a dominant emission band centered at 700 nm with a lifetime of 21.60 µs, ascribing to the delayed fluorescence of the methoxytetraphenyl-vinylpyridine part stimulated by PRET.
To further confirm the multivalent cascade confined single-molecular PRET luminescence behavior, a series of control experiments have been performed. The reference molecule PY-1 displayed CB[8]-induced strong RTP emission around 520 nm with a lifetime of 388.20 µs (Supplementary Fig. 33), suggesting that the phosphorescence emission of the TPE-DPY assembly emanated from the phenylpyridine units. The steady-state PL spectrum of TPE-1/CB[7] excited by 450 nm presented an emission peak at 720 nm (Fig. 5d) with a nanosecond lifetime measured as 2.74 ns, revealing the pure fluorescence properties of methoxy tetraphenylvinylpyridine unit (Supplementary Fig. 34). Similarly, under the excitation of 450 nm, TPE-DPY, TPE-DPY/CB[7]/CB[8] and TPE-DPY/CB[7]/CB[8]@HACD showed a fluorescence emission of 720 nm, where the lifetime was measured as 0.71 ns, 0.74 ns and 0.85 ns, respectively. No emission signal was obtained in the delay spectrum, verifying the property of HACD-mediated NIR delayed fluorescence at 700 nm that was excited by 333 nm (Supplementary Fig. 35). Furthermore, a large overlap was captured between the absorption spectra of TPE-1/CB[7] and the phosphorescence spectra of PY-1/CB[8] (Fig. 5d), which provided a prerequisite for the PRET derived from phenyl pyridines unit to methoxy-tetraphenylethylene portion within a single-molecule. It is worth noting that no effective PRET phenomenon was obtained in the HACD-assembly doping system with PY-1/CB[8] as the donor and TPE-1/CB[7] as the acceptor, which may be caused by the uncontrollable large distance between the donor and the acceptor without covalent connection, hindering the generation of the energy transfer process (Supplementary Fig. 36). Moreover, the addition of bare HA into the TPE-DPY/CB[7]/CB[8] solution cannot activate an efficient PRET like HACD, highlighting the important role of β-CD in this cascade assembly process (Supplementary Fig. 37). Accordingly, the binding behavior between the reference TPE-1 and β-CD was investigated by 1H NMR spectra. It was shown that upon the addition of β-CD, the protons on methoxyphenyl (H1, H2) in TPE-1 shifted slightly to low-field, while the protons in styryl pyridiniums remained unchanged (Supplementary Fig. S38), indicating the complexation of β-CD and methoxyphenyl unit. and the association constant of TPE-1/β-CD was determined to be 344.09 M− 1 (Supplementary Fig. 39). These experimental results consistently indicated that the anion effect of HA and the encapsulation of β-CD to methoxy-tetraphenylethylene moiety contributed to the reconstruction of topological morphology for TPE-DPY/CB[7]/CB[8], which facilitated the single intramolecular PRET process leading to 700 nm NIR delayed fluorescence emission with large Stokes shift of 367 nm (Fig. 5c). Commonly, the phosphorescence spectra of the assemblies TPE-DPY/CB[7]@HACD and TPE-DPY/CB[8]@HACD also exhibited NIR emission peaks at 700 nm, implying the universality of HACD activation for single-molecular PRET process (Supplementary Fig. 40).
In order to explore the application of macrocyclic confinement and HACD-activated single-molecule PRET system, cell imaging experiments were constructed. First, Human cervical carcinoma cells (Hela cells) and human embryonic kidney cells (293T cells) were treated with TPE-DPY/CB[7]/CB[8]@HACD for 12h, respectively, and then incubated with Hoechst and Mito-Tracker Green for localization experiment. The confocal laser scanning microscopy (CLSM) experiments were performed to investigate the intracellular NIR emission signals. As shown in Figs. 6a and 6c, Hela cells exhibited a bright NIR luminescence in a red channel (650–750 nm), whereas almost no red emission signal was found for normal 293T cells. These imaging results implied that TPE-DPY/CB[7]/CB[8]@HACD was preferentially internalized by cancer cells rather than normal cells, which may be caused by the overexpressed HA receptors for cancer cells. Furthermore, colocalization analysis demonstrated that the NIR luminescence signal overlapped well with the green signal of Mito Tracker, which corresponded to the yellow region in the merged image (Fig. 6b). It revealed the ability of TPE-DPY/CB[7]/CB[8]@HACD for targeted mitochondria imaging in cancer cells, and the high Pearson correlation provided strong evidence for this result (Supplementary Fig. 41). Finally, CCK-8 assays were conducted to evaluate cytotoxicity experiments on the above two cells, and the high survival rate indicated low cytotoxicity of the assembly TPE-DPY/CB[7]/CB[8]@HACD (Supplementary Fig. 42).