Isomeric double-cavity coordination cages: to encapsulate or not to encapsulate the guest

Abstract

A series of three isomeric double-cavity discrete coordination cages with Pd₃L₄ formulation, constructed from Pd(II) and regioisomeric tris-monodentate ligands, are compared for binding pyrazine N,N′-dioxide. Two of the isomeric cages efficiently encapsulated the guest in both cavities, whereas one of the isomers rejected the guest completely.

---

The well-known Pd₂L₄-type single-cavity discrete coordination cages (SCDCCs) containing one 3-D cavity,^1–6 and relatively new Pd₃L₄-type multi-cavity discrete coordination cages (MCDCCs) containing two 3-D cavities^4–8 are typically prepared by complexation of Pd(II) with a suitable bis-monodentate ligand and tris-monodentate ligand, respectively (Fig. 1(A) and 2(A)). The framework of a Pd₃L₄-type MCDCC is composed of two Pd₂L₄-type sub-frameworks that are conjoined at a common metal centre.^4–8 Such Pd₂L₄-type and Pd₃L₄-type cages are usually symmetrical, as the chosen bis-/tris-monodentate ligands possess C₂-symmetry.

Fig. 1 Synthesis and guest binding behaviours of isomeric single-cavity discrete coordination cages 1·4NO₃, 2·4NO₃ and 3·4NO₃ prepared by combining Pd(NO₃)₂ with (A) L1 (see ref. 12), (B) L2 and (C) L3; counteranions are not shown in the cartoon diagrams. Crystal structure of (PZDO)⊂1·4NO₃ (see ref. 12) and (PZDO)⊂2·4OTs (this work) are given for comparison; counteranions and solvents are not shown for clarity.

To lower the symmetry of the cages, unsymmetrical ligands have been employed recently.^3–10 The Pd₂L₄ type of cage prepared using an unsymmetrical bis-monodentate ligand may result in four possible orientational isomers (Fig. 1(C)), but the selective formation of one of the isomers could be achieved when geometrical complementarity is encoded in the ligand design.^4–6 Similarly, unsymmetrical tris-monodentate ligands may afford a mixture of isomeric Pd₃L₄-type cages but preparation of a single isomer is possible by suitable ligand design.^11–13 Alternatively, the symmetry in the cages can also be lowered by incorporating more than one type of designer ligand in the framework of the cage molecules, e.g. Pd₂L^a₂L^b₂ and Pd₃L^a₂L^b₂.^3,13 However, we focus here on cages prepared using only one type of ligand.

The term “ligand isomerism in coordination cages” embraces isomeric coordination cages. A family of such isomeric cages can be prepared individually in separate pots by combining a metal component with corresponding regioisomeric ligands.^13–19 We are interested in exploring ligand isomerism in Pd(II)-based SCDCCs and MCDCCs.

In this work, we compared the guest binding abilities of a series of three isomeric Pd₂L₄-type SCDCCs and a related series of three isomeric Pd₃L₄-type MCDCCs equipped with a single and double cavity, respectively. Unusual behaviour was observed for the binding of pyrazine N,N′-dioxide (PZDO) by the cages. While two of the isomeric cages of each series could encapsulate PZDO, the other isomeric cages rejected PZDO.

The regioisomeric bis-monodentate ligands L1,¹²L2²⁰ and L3 explored in this work (Fig. 1(A)) possess two terminal pyridine-3-yl units as donor sites. Each of the donor sites are connected with a central meta-phenylene spacer via amide linkages; thus, the ligands are incorporated with two amide groups. A closer comparison of the structures would show that both N-amide atoms are connected with the phenylene spacer at meta-positions, and the C-carbonyl atoms are linked at the meta-positions of terminal pyridyl groups in symmetrical ligand L1. Compared to L1, the connectivity for both amide moieties is upturned, as can be seen in symmetrical ligand L2 (Fig. 1(B)). Such connectivity is upturned for only one of the amide moieties, as in unsymmetrical ligand L3 (Fig. 1(C)).

The regioisomeric tris-monodentate ligands L4,²¹L5, and L6 (Fig. 2) possess two terminal pyridine-3-yl units and one internal pyridine-3,5-diyl unit as donors. All the ligands have C₂-symmetry along the 3,5-disubstituted internal pyridine. The bis-monodentate fragments of the tris-monodentate ligands, i.e. a structure spanning from the internal pyridine to any one of the terminal pyridines, also possess C₂-symmetry along the meta-phenylene spacer in the case of ligands L4 (Fig. 2(A)) and L5 (Fig. 2(B)). This can be visualized by comparing a fragment of L4 with symmetrical L1, and likewise comparing a fragment of L5 with symmetrical L2. However, the C₂-symmetry of the above-mentioned bis-monodentate fragments is lost in ligand L6 (Fig. 2(C)) (compare a fragment of L6 with an unsymmetrical L3).

Fig. 2 Synthesis and guest binding behaviours of isomeric double-cavity discrete coordination cages 4·6NO₃, 5·6NO₃ and 6·6NO₃ prepared by combining Pd(NO₃)₂ with (A) L4 (see ref. 21), (B) L5 and (C) L6; counteranions are not shown in the cartoon diagrams. Crystal structure of (PZDO)₂⊂4·6NO₃ (see ref. 21) (PZDO)₂⊂5·6OTs (this work) and 6·6NO₃ (this work) are given for comparison; counteranions and solvents are not shown for clarity.

Ligands L1, L2 and L4 have been reported previously,^12,20,21 whereas ligands L3, L5 and L6 are new. Ligand L2 was synthesized by modifying a procedure in the literature,²⁰ whereas new ligands L3, L5 and L6 were obtained through multistep processes (SI, Schemes S1–S4). The synthesized ligands and precursors were thoroughly characterized by 1D/2D NMR spectroscopy, along with the ESI-MS technique (SI, Fig. S1–S38). The Pd(II) complexes prepared from known ligands L1 and L4 and their guest (PZDO) binding behaviours (Fig. 1(A) and 2(A)), required here for comparison, are already available in the literature.^12,21 However, L1 was prepared for performing some additional experiments, while synthesis of L4 was not required.

Complexation of Pd(NO₃)₂ with ligands L1–L6 in DMSO medium produced the cages [Pd₂(L1)₄](NO₃)₄, 1·4NO₃; [Pd₂(L2)₄](NO₃)₄, 2·4NO₃; [Pd₂(L3)₄](NO₃)₄, 3·4NO₃; [Pd₃(L4)₄](NO₃)₆, 4·6NO₃; [Pd₃(L5)₄](NO₃)₆, 5·6NO₃; and [Pd₃(L6)₄](NO₃)₆, 6·6NO₃. Among these cages, only 1·4NO₃ and 4·6NO₃ were reported earlier.^12,21 The structures of all the cages are depicted in cartoon format only (Fig. 1 and 2), but counteranions are not shown. Typically, the concentration of Pd(NO₃)₂ is maintained at 10 mM, unless there is a need for a special experiment.

The single-cavity-containing binuclear Pd₂L₄-type cage 1·4NO₃ remained in dynamic equilibrium with minor amounts of trinuclear Pd₃L₆-type cage [Pd₃(L1)₆](NO₃)₆, 1′·6NO₃.¹² The combination of a required amount of PZDO (5 to 6 mM in final solution) with the equilibrium mixture that is 10 mM in Pd(II), generated an amount of [(PZDO)⊂Pd₂(L1)₄](NO₃)₄, (PZDO)⊂1·4NO₃.¹² In the presence of 5–6 mM of PZDO, no further changes were observed by letting the solution stand for as long as 24 h (this work: SI, Fig. S59). However, upon the addition of an excess amount of PZDO (25 mM in the final solution), a guest-induced equilibrium shift was also observed that afforded (PZDO)⊂1·4NO₃ exclusively, as reported by our group.¹²

In the present work, we prepared and found that the binuclear cage 2·4NO₃ also remained in equilibrium with minor amounts of trinuclear cage [Pd₃(L2)₆](NO₃)₆, 2′·6NO₃ (SI, Scheme S5). Characterization by ¹H NMR, 2D-COSY, 2D-NOESY (SI, Fig. S39–S41), and ESI-MS (SI, Fig. S51) confirmed the presence of both binuclear and trinuclear assemblies. The ¹H NMR spectrum at room temperature showed two sets of new signals compared to free L2 (Fig. 3(i) and (ii)) corresponding to the bi/trinuclear complexes. The proportion of binuclear 2·4NO₃ was found to have increased upon lowering the concentration or increasing the temperature or letting the mixture stand at room temperature for a longer time (SI, Fig. S55–S57).

Fig. 3 Partial ¹H NMR (500 MHz, 298 K, DMSO-d₆) spectra of (i) ligand L2; (ii) mixture of 2·4NO₃ and 2^′·6NO₃; (iii) (PZDO)⊂2·4NO₃; (iv) ligand L3; (v)/(vi) mixture of 3·4NO₃ and 3′·6NO₃ before and after the addition of excess PZDO; (vii) free PZDO.

The combination of a required amount of PZDO (5 to 6 mM) with the mixture of bi/trinuclear complexes at room temperature resulted in [(PZDO)⊂Pd₂(L2)₄](NO₃)₄, (PZDO)⊂2·4NO₃ and the disappearance of binuclear 2·4NO₃ while an amount of trinuclear 2′·6NO₃ remained in solution when monitored at 4 h by recording the ¹H NMR spectra. Thereafter, the trinuclear species also disappeared either upon letting the solution stand for 24 h (SI, Fig. S62), unlike the case of 1·4NO₃, or upon the addition of an excess amount of PZDO (SI, Fig. S63), similar to the case of 1·4NO₃. The formation of (PZDO)⊂2·4NO₃ was confirmed by ¹H NMR, ESI-MS and single crystal structure (Fig. 1(B), 3(ii), (iii) and SI, Fig. S60–S63). The above-described result indicated the superiority of 2·4NO₃ for binding PZDO compared to isomeric 1·4NO₃.

A mixture of products resulted due to complexation of Pd(NO₃)₂ with L3, as suggested by the multiple sets of peaks in the ¹H NMR spectrum (Fig. 3(v)). The unsymmetrical nature of ligand L3 should result in four isomers of [Pd₂(L3)₄](NO₃)₄, 3·4NO₃ formulation (Fig. 1C). Although a binuclear complex was detected in the ESI-MS (SI, Fig. S52), no trinuclear complexes were observed; however, we cannot rule out the copresence of an isomeric mixture of trinuclear [Pd₃(L3)₆](NO₃)₆, 3′·6NO₃. The addition of PZDO, even in an excess amount, to the isomeric mixture showed only marginal changes in the ¹H NMR spectrum, indicating negligible or no guest encapsulation by 3·4NO₃ (Fig. 1(C), 3(vi) and SI, Fig. S64). Thus, among the isomeric SCDCCs, 1·4NO₃, 2·4NO₃ and 3·4NO₃, the former two cages encapsulated PZDO and the last probably rejected PZDO.

The double-cavity trinuclear Pd₃L₄-type cage 4·6NO₃ prepared from Pd(NO₃)₂ and L4 is capable of encapsulating two units of PZDO, one in each cavity, forming (PZDO)₂⊂4·6NO₃ (Fig. 2(A)).²¹

Complexation of Pd(NO₃)₂ with L5 and L6 separately in a 3 : 4 ratio in DMSO-d₆ yielded discrete Pd₃L₄ assemblies [(NO₃)_2x⊂Pd₃(L5)₄](NO₃)_6–2x, 5·6NO₃; and [(NO₃)_2x⊂Pd₃(L6)₄](NO₃)_6–2x, 6·6NO₃, respectively (Fig. 2(B) and (C)). The ¹H NMR spectra showed complexation-induced single sets of signals for each case, confirming single discrete products (Fig. 4 and SI Fig. S45–S50), whereas ESI-MS study confirmed Pd₃L₄ formation (SI, Fig. S53 and S54).

Fig. 4 Partial ¹H NMR (500 MHz, 298 K, DMSO-d₆) spectra of (i) ligand L5; (ii) 5·6NO₃; (iii) (PZDO)₂⊂5·6NO₃; (iv) ligand L6; (v) 6·6NO₃; (vi) a mixture of 6·6NO₃ and PZDO (excess) showing no encapsulation of PZDO; (vii) free PZDO.

Any one of the cavities of the isomeric MCDCCs, 4·6NO₃, 5·6NO₃ and 6·6NO₃, resembles the cavity of the isomeric SCDCCs, 1·4NO₃, 2·4NO₃, and (4,0),(0,4)-3·4NO₃, respectively. The distinct chemical environment provided by the inner spaces of any one of the isomers of the MCDCCs, compared to the other two isomers, is expected to display varying degrees of response to a common guest like PZDO, in line with the differences observed for guest binding by the SCDCCs.

Cage 4·6NO₃ encapsulates PZDO, but shows considerable broadening of ¹H NMR signals where the signals of bound guests are not observed, although ESI-MS and crystal structures supported the formation of (PZDO)₂⊂4·6NO₃.²¹ The guest binding behaviours of other isomeric MCDCCs were studied in this work by the gradual addition of PZDO. The addition of two equiv. of PZDO to 5·6NO₃, prepared in DMSO-d₆, resulted in the formation of (PZDO)₂⊂5·6NO₃, as confirmed by a sharp ¹H NMR signal with characteristic downfield shifts for α-pyridyl protons, H(a₅) (Δδ ∼ 0.8 ppm) and H(k₅) (Δδ ∼ 0.7 ppm) (Fig. 4(iii) and SI Scheme S11). Two peaks at ∼10.2 and 10.3 ppm indicated signals corresponding to encapsulated PZDO, which also suggested that the bound PZDO may or may not rotate around the “Pd_terminal–(O_PZDO⋯O_PZDO)–Pd_internal” axis, but the positions of PZDO oxygen atoms are not exchanged by vertical flipping of bound PZDO. Titration was performed by adding 0.2–2.0 equiv. of PZDO per MCDCC to a solution of 5·6NO₃ (SI, Fig. S65). Upon addition of less than 2 equiv. of PZDO, the ¹H NMR spectrum showed signals for native 5·6NO₃, singly occupied host–guest species (PZDO)⊂5·6NO₃ and doubly occupied host–guest species (PZDO)₂⊂5·6NO₃ (SI Fig. S66). The ESI-MS (SI Fig. S67) study supported the existence of 5·6NO₃ and (PZDO)⊂5·6NO₃, but (PZDO)₂⊂5·6NO₃ could not be detected. The mixture was then monitored by recording ¹H NMR (SI Fig. S66), which showed no further changes even after 72 hours; hence the cage 5·6NO₃ appears to show slow exchange host–guest dynamics.²² As guest binding was completed upon the addition of approximately two equiv. of PZDO a high binding affinity (K_a > 10⁵ M⁻¹) is proposed.²²

The low-symmetry MCDCC 6·6NO₃ showed no PZDO binding even after the addition of an excess amount of PZDO, as the ¹H NMR showed no changes in the position of signals (Fig. 4(vi), SI Scheme S12 and Fig. S68). This is in line with the proposed rejection of the guest by the related SCDCC 3·4NO₃.

The NOE analysis of the cage 6·6NO₃ (SI, Fig. S50) revealed that all amide NH groups closer to the terminal pyridine units are oriented outward, as the NH protons showed cross peaks with relevant protons located only at the outer surface of the cage. Similarly, the amide NH groups closer to the internal pyridine units are directed inward, as the NH protons showed cross peaks with relevant protons located only at the inner surface of the cage. In contrast, NOE analysis of the other isomeric cages 4·6NO₃ and 5·6NO₃ (ref. 21, and SI, Fig. S47) indicates conformational flexibility, as evidenced by multiple NOE cross peaks for the amide NH protons with the protons located both at the outer and inner surfaces of the cages. We speculate that the structural flexibility might facilitate guest binding and the rigidity might restrict the guest binding in this variety of host molecules. However, such an observation may not be a guiding principle.

Complex 6·6NO₃ was also characterized by single crystal X-ray diffraction analysis (Fig. 2(C) and SI section S7). Single crystals of [((DMF)_0.84(DMSO)_0.16)₂(NO₃)₂⊂Pd₃(L6)₄](NO₃)₄ were obtained by slow diffusion of THF vapor into a solution of 6·6NO₃ that was prepared in DMF. Two units of nitrate ions are encapsulated, one in each cavity; also, one nitrate was located near the central pyridine around the outer periphery of the cage. Additionally, one of the DMF molecules lies inside the cavity of the cage and exhibits substitutional disorder, wherein a DMSO molecule is also observed at the same location; the DMF : DMSO ratio being 0.84 : 0.16.

Single crystals of host–guest complex [(PZDO)⊂Pd₂(L2)₄](OTs)₄, (PZDO)⊂2·4OTS were obtained by slow vapor diffusion of dioxane : THF (1 : 1) into its DMSO solution. The structure of (PZDO)⊂2·4OTs revealed a slightly tilted PZDO within the Pd₂L₄ framework (Fig. 1(B)). Hydrogen bonding interaction was evident from C–H(host)⋯O–N(guest) (∼2.4 Å) and C=O(Host)⋯H–C(guest) (2.1–2.3 Å) contacts, while electrostatic Pd⋯O interactions were indicated by a separation of ∼2.8 Å. Similarly, X-ray diffraction quality crystals of [(PZDO)₂⊂Pd₃(L5)₄](OTs)₆, (PZDO)₂⊂5·6OTs were obtained by diffusing benzene : THF (1 : 1) vapor into its DMSO solution. The structure showed a Pd₃L₄ complex with two encapsulated PZDO, slightly tilted, as observed in SCDCC. The structure exhibited similar ligand conformation, hydrogen bonding, and electrostatic interactions as in guest-encapsulated SCDCC, (PZDO)⊂2·4OTS (Fig. 2(C), SI section S7 for details).

In summary, we probed a series of isomeric Pd₃L₄-type double-decker MCDCCs prepared by combining Pd(II) with tris-monodentate regioisomeric ligands. A guest binding study using PZDO as a guest and isomeric cages as hosts showed differential guest uptake abilities, ranging from efficient binding to rejection of the guest. Overall, this work highlights the potential of ligand isomerism as a parameter responsible for emerging supramolecular behaviour.

Conflicts of interest

There are no conflicts to declare.

Data availability

All relevant data have been included in the manuscript and supplementary information (SI). Supplementary information: synthesis and characterization of the compounds (ligands, cages, PZDO encapsulated cages), and crystal structure data. See DOI: https://doi.org/10.1039/d5dt02147b.

CCDC 2452518–2452520 contain the supplementary crystallographic data for this paper.^23a–c

Acknowledgements

D. K. C. thanks the Anusandhan National Research Foundation (ANRF) (formerly Science and Engineering Research Board (SERB)), Government of India (project no. CRG/2022/004413), for financial support. D. K. C. thanks IIT Madras for financial support through a Center under Institute of Eminence program (IoE center of Molecular Architecture). We thank SAIF, IIT Madras for single crystal X-ray diffraction facility. We also thank the CoE on Molecular Materials and Functions and IoE Center of Molecular Architecture for IoE funded single crystal X-ray diffraction facility (6·6NO₃); we appreciate Ayan Ghosh for rendering help in data collection. We thank the Department of Chemistry, IIT Madras for NMR facility and DST-FIST funded ESI-MS facility. We thank the P. G. Senapathy Centre for Computing Resources, IIT Madras, for providing access to Gaussian 16 package. We dedicate this communication to Prof. N. N. Murthy on the occassion of his superannuation.

References

1. R. Saha, B. Mondal and P. S. Mukherjee, Chem. Rev., 2022, 122, 12244–12307.
2. M. Yoshizawa and L. Catti, Acc. Chem. Res., 2019, 52, 2392–2404.
3. L. Neukirch and G. H. Clever, Chem. Sci., 2025, 16, 12242–12276.
4. J. E. M. Lewis, Chem. Commun., 2022, 58, 13873–13886.
5. S. Sharma, M. Sarkar and D. K. Chand, Chem. Commun., 2023, 59, 535–554.
6. C. T. McTernan, J. A. Davies and J. R. Nitschke, Chem. Rev., 2022, 122, 10393–10437.
7. A. C. Pearcy and J. D. Crowley, Chem. – Eur. J., 2023, 29, e202203752.
8. R. A. S. Vasdev, D. Preston and J. D. Crowley, Chem. – Asian J., 2017, 12, 2513–2523.
9. J. E. M. Lewis and J. D. Crowley, ChemPlusChem, 2020, 85, 815–827.
10. L. K. Moree, L. A. V. Faulkner and J. D. Crowley, Chem. Soc. Rev., 2024, 53, 25–46.
11. J. E. M. Lewis, Angew. Chem., Int. Ed., 2022, 61, e202212392.
12. S. S. Mishra, S. Krishnaswamy and D. K. Chand, J. Am. Chem. Soc., 2024, 146, 4473–4488.
13. A. Kumar, S. Krishnaswamy and D. K. Chand, Angew. Chem., Int. Ed., 2025, 64, e202416332.
14. D. Preston, S. M. McNeill, J. E. M. Lewis, G. I. Giles and J. D. Crowley, Dalton Trans., 2016, 45, 8050–8060.
15. H. Dasary, R. Jagan and D. K. Chand, Inorg. Chem., 2018, 57, 12222–12231.
16. H.-J. Yu, Z.-M. Liu, M. Pan, K. Wu, Z.-W. Wei, Y.-W. Xu, Y.-N. Fan, H.-P. Wang and C.-Y. Su, Eur. J. Inorg. Chem., 2018, 80–85.
17. S. Samantray, S. Bandi and D. K. Chand, Beilstein J. Org. Chem., 2019, 15, 1129–1140.
18. H. Dasary, M. Sarkar and D. K. Chand, Chem. Commun., 2022, 58, 8480–8483.
19. H. B. Gearing, T. Sohnel, P. Young, L. Lisboa, L. J. Wright, J. D. Crowley and C. G. Hartinger, Chem. Commun., 2024, 60, 10950–10953.
20. N. L. S. Yue, Z. Qin, M. C. Jennings, D. J. Eisler and R. J. Puddephatt, Inorg. Chem. Commun., 2003, 6, 1269–1271.
21. T. Dakua, S. S. Mishra, A. Kumar, S. Krishnaswamy and D. K. Chand, Chem. Commun., 2025, 61, 8007–8010.
22. P. Thordarson, Chem. Soc. Rev., 2011, 40, 1305–1323.
23. (a) CCDC 2452518: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2nb1hk; (b) CCDC 2452519: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2nb1jl; (c) CCDC 2452520: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2nb1km.

---