Size-selective formation of metallonanobelts via tethering-template-directed self-assembly

Abstract

We propose a novel rational strategy for template-directed self-assembly using “tethering templates” that have common interaction units and a distinct sized core. By employing two kinds of tethering templates bearing common triethylene glycol chains and differently sized pillar-shaped cores, pillar[n]arenes (n = 5, 6), the tetra- and pentanuclear metallomacrocycles were selectively obtained during the self-assembly of 2,3,6,7-tetraaminotriptycene and the Pd²⁺ ions, under thermodynamic control.

---

Coordination-driven self-assembly of organic ligands and metal ions is a key strategy for constructing ordered, discrete structures with functionalities such as chemical separation,¹ storage² and catalysis.³ Among the various self-assembled structures obtained via this approach, metallomacrocycles have attracted particular interest as versatile building blocks for functional materials, including molecular receptors⁴ and one-dimensional tubular structures.⁵ However, self-assembly processes for metallomacrocycles often lead to a mixture of different ring sizes since their thermodynamic stabilities are close to each other.⁶ To selectively obtain a single product, template-directed self-assembly is a useful strategy because templates tightly bind to the desired structure via hydrogen bonding, electrostatic interactions, or van der Waals interactions, thus stabilizing the target metallomacrocycles relative to other assembled structures. To construct different sizes of metallomacrocycles, distinct templates are required for each structure. While inorganic ions⁷ and small neutral organic molecules⁸ have been commonly used as templates, they are inherently limited in size and structural tunability (Fig. 1a). Moreover, the strength and nature of their intermolecular interactions depend on their atomic composition and molecular geometry. Thus, the rational design of templates capable of selectively directing the formation of a desired structure remains a formidable challenge. The development of template molecules with a greater structural tunability, particularly those tunable in both size and interaction sites, provides an effective approach to overcoming these limitations.

Fig. 1 Schematic representations of the formation of metallomacrocycles by self-assembly of organic ligands and metal ions using conventional templates (a) and tethering templates (b).

We have recently reported template-directed self-assembly based on unique interactions during the formation of a belt-shaped metallomacrocycle, referred to as a “metallonanobelt”, via the self-assembly of 2,3,6,7-tetraaminotriptycene (L) and the Pd²⁺ ion (Fig. 2a(i)).⁹ We found that the complexation of L with the Pd²⁺ ion in the presence of T-P6, a pillar[6]arene derivative having triethylene glycol (TEG) chains, led to the selective formation of pentanuclear metallonanobelts (pentamer; Fig. 2a(ii)). This template effect is attributed to both the multiple hydrogen-bonding interactions between the TEG chains and the [Pd-(o-phenylenediamine)₂] units¹⁰ in the pentamer and the size-exclusion by the pillar[6]arene scaffold. Based on this finding, we now propose a more general and modular strategy involving the “tethering templates”, which integrate a chemically tunable size-exclusion core and common interaction sites (Fig. 1b), allowing rational template-directed self-assembly for the formation of the desired sized metallomacrocycle. In this study, we applied this tethering template strategy for the formation of metallonanobelts. We designed and synthesized various tethering templates bearing oligoethylene glycol (OEG) chains as common interaction sites (Fig. 2b), and successfully achieved the selective formation of a tetranuclear metallonanobelt (tetramer) using a pillar[5]arene derivative as a template (Fig. 2a(iii)).

Fig. 2 (a) Schematic representations of the self-assembly of L and the Pd²⁺ ion without templates (i), with T-P6 (ii) and with T-P5 (iii). In the absence of templates, a mixture of metallonanobelts with different sizes can be obtained. In contrast, two distinct-sized metallonanobelts can be selectively formed in the presence of appropriate templates. (b) Chemical structures of the tethering templates.

Although our previous study suggested that multiple hydrogen bonds played an important role for the template effect, the detailed interactions at the molecular level remained elusive. To address this, we performed an X-ray crystallographic analysis on a single crystal of the host–guest complex, a pentamer with T-P6 (Fig. 3a, b and Table S1). In the crystal structure, two crystallographically distinct host–guest complexes were identified (Fig. S1). In both structures, the highly-distorted pillar[6]arene core of T-P6 is situated near the center of the pentamer cavity, and no π–π interaction was observed between the aromatic panels of T-P6 and the inner walls of the pentamer within the crystal lattice. In contrast, four or five of the twelve TEG chains of T-P6 tightly entwined with the [Pd-(o-phenylenediamine)₂] units, forming multiple hydrogen bonds between NH₂ groups of the [Pd-(o-phenylenediamine)₂] units and oxygen atoms of the TEG chains (Fig. 3c). These results indicate that multiple hydrogen bonding interactions are the main factor for stabilizing the host–guest complex.

Fig. 3 (a) Space-filling and (b) ball-and-stick models for the crystal structure of the pentamer·T-P6 complex. The P6 cores and TEG chains are colored pale green and pink, respectively, while the oxygen atoms of the TEG chains are colored red. The solvent molecules, triflate anions and disorder are omitted for clarity. (c) Magnified views of the hydrogen bonding interactions between the NH₂ groups and TEG chains. The N–H⋯O hydrogen-bonding is indicated by the blue dotted lines. Hydrogen atoms, except those in the NH₂ groups, are omitted for clarity.

The results obtained from the crystallographic analysis motivated us to attempt various kinds of tethering templates having common OEG chains as common interaction units and differently sized cores for the formation of different sizes of metallonanobelts. We designed ten different tethering templates based on the pillar[6]arene (P6), pillar[5]arene (P5), fullerene (C₆₀), pyrogallol[4]arene (Py4) and cyclotriveratrylene (CTV) as size-exclusion cores, each bearing OEG chains of varying chain lengths (Fig. 2b; synthesis and characterization, see the SI). Since these tethering templates, except for the series of P6, have smaller cores compared to the pillar[6]arene (Fig. S2), we expected that the smaller-sized metallonanobelt (i.e. tetramer or trimer) would be obtained by using them as a template. The influences of the length of the OEG chains on the strength of the host–guest interaction and the template effect were also investigated.

Prior to exploring their templating effect, we evaluated the host–guest interaction between the synthesized templates and the isolated pentamer in acetonitrile at 25 °C using ¹H NMR titration and isothermal titration calorimetry (ITC) (Table 1). We first compared the binding constants for a series of P6-based tethering templates in order to elucidate how the OEG chain length affects the host–guest interactions. The K_a values for Te-P6, T-P6 and D-P6 were determined to be 3.0 × 10⁶, 1.9 × 10⁶ and 1.4 × 10³ M⁻¹, respectively, indicating that longer OEG chains enhance the binding affinity (Fig. S3–S5). A similar trend was observed for the P5- and C₆₀-based tethering templates (Fig. S6–S10). The relative binding constants (K_r) of the tethering templates bearing tetraethylene glycol chains (Te-X: X = P6, P5, C₆₀) for those possessing TEG chains (T-X) were estimated to be 3.3 for the P5-series and 160 for the C₆₀-series, both higher than the K_r value of 1.6 observed for the P6-series. These differences likely arise from the smaller core sizes of P5 and C₆₀, which may limit the ability of shorter TEG chains to form multiple hydrogen bonds with the [Pd(o-phenylenediamine)₂] units. Comparing T-CTV1 and T-CTV2, which both have the CTV scaffold and the TEG chains but differ only in the number of TEG chains, the relative binding constant was found to be 100 (Fig. S11 and S12). Moreover, T-Py4, which has a similar bowl-shaped scaffold but has twelve TEG chains, exhibited an approximately 13-fold greater K_a value than T-CTV2 (Fig. S13). These results highlight that not only the length but also the number of OEG chains plays a critical role in the binding affinity.

Table 1 Binding constants (K_a) for the pentamer and molar ratios of three metallonanobelts upon self-assembly of L and the Pd²⁺ ion in the absence or presence of tethering templates

Tethering template^a	K a (M^−1) for the pentamer	Molar ratios of pentamer : tetramer : trimer^d
a Values in brackets indicate the number of OEG chains. b Determined by fitting the changes in the chemical shifts of the pentamer observed in the ^1H NMR (0.5 mM based on pentamer, CD3CN) at 25 °C. c Determined by ITC experiments (0.1 mM based on pentamer, CH3CN) at 25 °C. d Estimated by integrating the proton signals corresponding to each metallonanobelt obtained from the self-assembly of L and the Pd^2+ ions in CD3CN at 50 °C, in the absence or presence of the tethering templates.
No template		24 : 54 : 22
D-P6 [12]	(1.4 ±; 0.3)^b × 10^3	21 : 56 : 23
T-P6 [12]	(1.9 ±; 0.1)^c × 10^6	100 : 0 : 0
Te-P6 [12]	(3.0 ±; 0.6)^c × 10^6	100 : 0 : 0
D-P5 [10]	(2.3 ±; 0.5)^b × 10^2	28 : 53 : 19
T-P5 [10]	(4.7 ±; 0.5)^c × 10^5	17 : 83 : 0
Te-P5 [10]	(1.6 ±; 0.2)^c × 10^6	35 : 65 : 0
T-Py4 [12]	(1.9 ±; 0.2)^c × 10^6	68 : 30 : 2
T-CTV1 [3]	(1.6 ±; 0.1)^b × 10^3	30 : 47 : 23
T-CTV2 [6]	(1.6 ±; 0.1)^c × 10^5	47 : 42 : 11
T-C60 [12]	(8.1 ±; 0.9)^b × 10^2	21 : 54 : 25
Te-C60 [12]	(1.3 ±; 0.1)^c × 10^5	43 : 40 : 17

---

We next investigated the template effect of a series of tethering templates. To selectively obtain the desired product by template-directed self-assembly, the template should interact more strongly with the product compared to the other sized assemblies. To evaluate the template effects, we compared the product distributions obtained in the presence and absence of each template under thermodynamic equilibrium. As a reference, the product distribution under template-free conditions ([L] = [Pd²⁺] = 4.0 mM in CD₃CN at 50 °C), which yielded a 24 : 54 : 22 molar ratio of the pentamer, tetramer and trimer metallonanobelt, was used (Table 1).

By exploring all the tethering templates, those based on the pillar[n]arenes showed a clear template effect on the self-assembly of L and the Pd²⁺ ion. Both T-P6 and Te-P6 selectively facilitated the pentamer formation, whereas D-P6 showed almost no template effect due to its weak binding interaction (Fig. S14). In contrast, the use of P5-based tethering templates resulted in a mixture of the pentamer and tetramer, with no trimer detected, suggesting a moderate template effect (Fig. 4a and Fig. S15). Although a complete selectivity was not achieved, the product distribution was more biased toward the tetramer than that under the template-free conditions. This selectivity is likely attributed to the smaller size-exclusion P5 core, which more closely fits the cavity of the tetramer. However, the P5-based tethering templates also retained a moderate affinity for the pentamer, leading to a mixture rather than an exclusive tetramer formation.

Fig. 4 ¹H NMR spectra (600 MHz, 25 °C) of the mixture of L, Pd²⁺ ion and T-P5 in CD₃CN after heating at 50 °C (a) and 70 °C (b) for 120 h. See Fig. 2a for signal assignments. (c) ¹H NMR spectrum (600 MHz, 25 °C) of the isolated tetramer in CD₃CN.

In contrast, non-pillar[n]arene-based templates showed little to no template effect, even with a relatively strong binding to the pentamer (Fig. S16–S18). This suggests that the binding strength alone does not account for the efficiency of the template effect. Thus, a strong binding strength only for the pentamer does not ensure selectivity if comparable interactions occur with other species. These tethering templates possess geometries that deviate from the belt-shaped cavities of the pentamer and are thus unlikely to engage in shape-complementary binding. Consequently, they are presumed to exhibit a similar binding strength for all metallonanobelts, which accounts for the reduced selectivity.

Based on the systematic experiments, P5-based tethering templates have emerged as promising candidates for the selective formation of the tetramer. Therefore, we aimed to increase the tetramer selectivity under the optimized conditions. Assuming that elevated temperatures would favor the entropically-preferred tetramer, we performed template-directed self-assembly at 70 °C using T-P5 and Te-P5. A ¹H NMR measurement revealed that the selectivity for the tetramer increased up to 92% for T-P5 and 82% for Te-P5 (Fig. 4b and Fig. S19). The resulting ¹H NMR spectra displayed complicated signals of the tetramer deviating from the expected D_4h symmetry, suggesting a pseudo C_2h symmetry in solution, likely due to structural deformation during the template inclusion. For the host–guest complex of the tetramer with T-P5, the template molecule was successfully removed by reprecipitation using a low-polar solvent (for details; see the SI). The isolated tetramer was characterized by mass spectrometry (Fig. S20) and ¹H NMR, which showed simple signals consistent with the expected D_4h symmetry (Fig. 4c). The isolated tetramer exhibited a strong binding toward T-P5, with a K_a of 1.7 × 10⁶ M⁻¹ as determined by ITC (Fig. S21). This K_a value is 3.5 times higher than that of the pentamer. This substantial difference in K_a explains why T-P5 serves as an effective template for tetramer formation.

In summary, the present study has demonstrated that the tethering templates containing a size-exclusion core with common interaction units act as effective templates for the self-assembly of metallonanobelts, leading to the selective formation of the distinct-sized metallonanobelts. We also successfully isolated a guest-free tetranuclear metallonanobelt, which showed a stronger binding affinity towards T-P5 than the pentanuclear metallonanobelt. Unlike the conventional strategy, the tethering template strategy imposes no intrinsic limitations on the size or shape of the template structure, offering a versatile platform for constructing a wide variety of assemblies. Thus this strategy would be widely applied not only to a series of metallonanobelts but also to spherical capsular assemblies based on the Pd-[o-(phenylenediamine)₂] units. We are currently exploring this possibility.

This work was supported by the JSPS KAKENHI JP20H04667, JP21H01948, JP25K22273 and JP24H02215 (Y. S.) in Transformative Research Areas (A) JP24A202 Integrated Science of Synthesis by Chemical Structure Reprogramming, and the JST FOREST (JPMJFR2329 to Y. S.). We also acknowledge financial support from The Asahi Glass Foundation, TEPCO Memorial Foundation, The Hibi Science Foundation, and World Premier International Research Center Initiative (WPI), MEXT, Japan.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d5cc04986e.

CCDC 2481995 contains the supplementary crystallographic data for this paper.¹¹

Notes and references

1. (a) D. Zhang, T. K. Ronson, Y.-Q. Zou and J. R. Nitschke, Nat. Rev. Chem., 2021, 5, 168–182; (b) Q.-H. Ling, Z.-C. Lou, L. Zhang, T. Jin, W.-T. Dou, H.-B. Yang and L. Xu, Chem. Soc. Rev., 2024, 53, 6042–6067.
2. (a) A. Galan and P. Ballester, Chem. Soc. Rev., 2016, 45, 1720–1737; (b) A. B. Grommet, M. Feller and R. Klajn, Nat. Nanotechnol., 2020, 15, 256–271.
3. (a) M. Yoshizawa, J. K. Klosterman and M. Fujita, Angew. Chem., Int. Ed., 2009, 48, 3418–3438; (b) C. J. Brown, F. D. Toste, R. G. Bergman and K. N. Raymond, Chem. Rev., 2015, 115, 3012–3035.
4. (a) T. Nakamura, Y. Kaneko, E. Nishibori and T. Nabeshima, Nat. Commun., 2017, 8, 129; (b) P. Howlader, B. Mondal, P. C. Purba, E. Zangrando and P. S. Mukherjee, J. Am. Chem. Soc., 2018, 140, 7952–7960.
5. (a) P. D. Frischmann, S. Guieu, R. Tabeshi and M. J. MacLachlan, J. Am. Chem. Soc., 2010, 132, 7668–7675; (b) K. Otsubo, Y. Wakabayashi, J. Ohara, S. Yamamoto, H. Matsuzaki, H. Okamoto, K. Nitta, T. Uruga and H. Kitagawa, Nat. Mater., 2011, 10, 291–295; (c) T. Fukino, H. Joo, Y. Hisada, M. Obana, H. Yamagishi, T. Hikima, M. Takata, N. Fujita and T. Aida, Science, 2014, 344, 499–504; (d) S. Kawano, A. Sasaki and K. Tanaka, Angew. Chem., Int. Ed., 2025, 64, e202422546.
6. (a) R. Chakrabarty, P. S. Mukherjee and P. J. Stang, Chem. Rev., 2011, 111, 6810–6918; (b) T. R. Cook and P. J. Stang, Chem. Rev., 2015, 115, 7001–7045.
7. (a) B. Hasenknopf, J.-M. Lehn, N. Boumediene, A. Dupont-Gervais, A. Van Dorsselaer, B. Kneisel and D. Fenske, J. Am. Chem. Soc., 1997, 119, 10956–10962; (b) R. L. Paul, Z. R. Bell, J. C. Jeffery, J. A. McCleverty and M. D. Ward, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 4883–4888; (c) C. S. Campos-Fernández, B. L. Schottel, H. T. Chifotides, J. K. Bera, J. Bacsa, J. M. Koomen, D. H. Russell and K. R. Dunbar, J. Am. Chem. Soc., 2005, 127, 12909–12923; (d) I. A. Riddell, M. M. J. Smulders, J. K. Clegg, Y. R. Hristova, B. Breiner, J. D. Thoburn and J. R. Nitschke, Nat. Chem., 2012, 4, 751–756; (e) S. Freye, R. Michel, D. Stalke, M. Pawliczek, H. Frauendorf and G. H. Clever, J. Am. Chem. Soc., 2013, 135, 8476–8479; (f) T. Zhang, L.-P. Zhou, X.-Q. Guo, L.-X. Cai and Q.-F. Sun, Nat. Commun., 2017, 8, 15898.
8. (a) M. Scherer, D. L. Caulder, D. W. Johnson and K. N. Raymond, Angew. Chem., Int. Ed., 1999, 38, 1587–1592; (b) K. Umemoto, K. Yamaguchi and M. Fujita, J. Am. Chem. Soc., 2000, 122, 7150–7151; (c) D. M. Wood, W. Meng, T. K. Ronson, A. R. Stefankiewicz, J. K. M. Sanders and J. R. Nitschke, Angew. Chem., Int. Ed., 2015, 54, 3988–3992; (d) K. Iizuka, H. Takezawa and M. Fujita, J. Am. Chem. Soc., 2024, 146, 32311–32316.
9. Y. Sakata, R. Yamamoto, D. Saito, Y. Tamura, K. Maruyama, T. Ogoshi and S. Akine, Inorg. Chem., 2018, 57, 15500–15506.
10. Y. Sakata, R. Nakamura, T. Hibi and S. Akine, Angew. Chem., Int. Ed., 2023, 62, e202217048.
11. CCDC 2481995: Experimental Crystal Structure Determination, 2025 DOI:10.5517/ccdc.csd.cc2p9qc3.

---

Footnote

† Dedicated to Professor Tatsuya Nabeshima on the occasion of his 70th birthday.

---