Tunable, shape-shifting capsule for dicarboxylates

Abstract

A cylindrical amide-based tricycle provides a ditopic framework for encapsulating both aliphatic and aromatic dicarboxylate anions. Due to the flexible spacer between the two macrocyclic receptor sites, it can modulate its shape to conform to dicarboxylates of varying lengths. In the uncomplexed form, the host is elongated along its cylindrical axis, but when encapsulating the guest, it compresses to ensure the best structural fit.

---

In 1978 Lehn and coworkers published the first example of an azacryptand anion host binding the multi-atomic linear ion, azide.¹ By elongating the cryptand viacyclophane spacers, the Lehn group later succeeded in obtaining the first crystallographic evidence for a cryptand-encapsulated pseudo-linear dicarboxylate ion.² Carboxylates are of course of great interest due to their roles in biological chemistry as well as in the environment.³ During the subsequent decades, the study of the binding of many different anions with supramolecular hosts has become a field of its own.⁴ Supramolecular chemistry has also provided the fuel for the development of working molecules, such as molecular devices and machines.⁵Hosts that can adjust their shapes upon binding, promoting chemical movement and transformations, are key targets for this fertile new area of chemistry. Examples of significant shape changing hosts include Badjić's translocating, gating molecular baskets,⁶ which function as a result of H-bond-dependent motions, and Sauvage's Cu(I) hermaphrodite double rotaxane,⁷ which stretches and shrinks depending on transition metal identity. Herein, we report a tunable molecular capsule that binds a number of aliphatic and aromatic dicarboxylates with high affinities by shifting its shape.

One of the main goals of our research has been to build supramolecular anion receptors of increasing dimensionality using simple amide/amine building blocks. The tricyclic cylinder 1 (Fig. 1) is the result of a number of years of research in our group exploring the effect of increasing dimensionality on anion binding.⁸ As part of our interest in this class of hosts, we synthesized the prototype tricyclic ligand, 2, which was found to bind both bifluoride and azide, the former ion with especially high affinity.⁹ As the next step in design, we decided to examine the effect of increasing cavity dimensions by lengthening the linkers, yielding the expanded cylindrical capsule 1. To our surprise 1 can readily tune its length to match certain dicarboxylates, as illustrated by the crystal structure results for the uncomplexed host, and the terephthalate and succinate complexes. In the free base form, the cylindrical host is elongated along the bridging axes, but when encapsulating guests, it compresses to the approximate length of the carboxylate. The tricycle may, therefore, provide a new prototype for contractable devices.

Fig. 1 Amide-based tricycle hosts 1 and 2.

By mixing dimethyl 2,6-pyridinedicarboxylate in a one-pot condensation reaction with tris(2-aminoethyl)amine (tren) in a 3 : 2 ratio, the tricycle 1 was obtained in 3% yield. The molecular capsule was characterized by mass spectrometry and NMR spectroscopy as well as single crystal X-ray diffractometry. Crystal structures of the uncomplexed host and complexes with two different dicarboxylates, succinate and terephthalate, were obtained.

The free base 1 crystallizes as the solvate 1·5EtOH·6H₂O in space groupC2/c and possesses crystallographic C_i symmetry. The overall structure consists of two tetraamidopyridine monocycles bridged by two diamidopyridine spacers, the resulting receptor somewhat resembling double-ended tongs (Fig. 2). The length of the spacers equals 10.051 Å (N1⋯N16a) (Table 1), which is much longer than that of the ethylene bridged tricycle 2 (∼3.8 Å).⁹ However the distance between the p-carbon atoms on the pyridines on the long axis (atoms e in Fig. 1) is over twice that distance, 21.962 Å. The two macrocycles are in a folded conformation, with the two pyridine rings of a given macrocycle separated by 3.725 Å. The distance between the bridgehead macrocyclic amines is 5.947 Å (N1⋯N16). The diamidopyridine groups of the two spacers between the macrocycles are staggered and orientate in an anti conformation. Interactions between a carbonyl oxygen atom on each bridge (O6 and O6a) and two of the nearest amido NH groups at 2.882 and 3.074 Å serve to lock in the configuration. Two EtOH molecules lie along the bridging diamides, but they are disordered and are not shown in Fig. 2.

Fig. 2 ORTEP at 30% ellipsoid probability (left) and space-filling (right) views of the free base 1. Only amide H atoms are shown in the ORTEP view on the left, and solvent molecules are omitted in both views for clarity.

Table 1 Selected distances [Å] defining the dimensions of the tricyclic ligand 1 for the free base, 1, the terephthalate, 1·tere, and succinate, 1·succ, complexes

Distance	1	1 ·tere	1 ·succ
Cylindrical length
N1–N16	5.947	9.162	8.691
N4–N19a	14.518	7.393	4.995
N13–N28a	13.213	4.830	3.853
Water pocket dimensions
N1–N16a	10.051	5.417	5.468
N16a–N42	3.274	3.171	2.962
N42–N33	4.565	4.511	4.783
N33–N1	3.787	3.128	3.009

---

With two macrocyclic binding pockets bridged by two flexible spacers, 1 is particularly suited for binding ditopic anions such as dicarboxylates as verified by the isolation of crystals of both the succinate and terephthalate complexes. The terephthalate complex, [n-Bu₄N]₂[1(terephthalate)]·4CH₃CN·6H₂O crystallizes in the P1̄ space group, again with crystallographic C_i symmetry. As seen in Fig. 3, the two macrocycles are flattened, and the terephthalate lies between them, inside the puffed-out pseudo-cylinder. With the new flattened shape, the N1⋯N16a distance between macrocycles has decreased sharply from 10.051 Å in the free base to 5.417 Å, while the N1⋯N16 separation across the macrocycles has increased from 5.947 Å to 9.162 Å – essentially inverted from the distances in the free base (Table 1). Now the p-carbon atoms of the two diamidopyridine bridges form the long axis, with C⋯C distances of 21.103 Å (atoms e′ in Fig. 1). The distance between the p-carbon atoms in the macrocyclic pyridines across the cylinder (atoms e in Fig. 1) is shortened to 7.282 Å. Hence, the tricyclic host has shifted its shape to incorporate the terephthalate guest. Each macrocycle forms a tetradentate chelate with a terminal carboxylate. Here the distances of note are those between the corresponding amides on opposing macrocycles that are H-bonded with the carboxylate oxygen atoms, N4⋯N19a and N13⋯N28a. These distances specify the “height” of the cylinder and are 7.393 and 4.830 Å, respectively, compared to 14.518 and 13.213 Å in the free base (Table 1). A fifth (apical) coordination site is occupied by a H₂O molecule that serves as a “cork” to hold the dicarboxylate in place. The two side diamidopyridine spacers provide extra binding pockets and accordingly hold two H₂O molecules, bound in a rather distorted four-fold H-bond coordination (Fig. 3). The other solvent molecules of crystallization do not lie within the host cavity.

Fig. 3 ORTEP at 30% ellipsoid probability (top) and space-filling (bottom) views of the terephthalate complex with 1. Only H-bonding H atoms are shown in the ORTEP view on the top, and the n-Bu₄N⁺ counterions and nonessential solvent molecules are omitted in both views for clarity.

The binding mode for succinate ion (Fig. 4) is similar to that of the terephthalate. The complex crystallizes as the mixed solvate, [Me₄N]₂[1(succinate)]·2CHCl₃·2CH₃CN·3H₂O in the space group P1̄. The dianion is entirely encapsulated once again, with each carboxylate group bound to the macrocyclic unit through four amide H-bonds. In this case, the puffed out pseudo-cylinder is even shorter, as measured from the N4⋯N19a and N13⋯N28a distances of 4.995 and 3.853 Å, respectively (Table 1). This allows for the shorter dicarboxylate to nestle better within its host (carboxylate C⋯C distances = 3.966 and 5.812 Å, for succinate and terephthalate, respectively). Instead of axial water molecules in the terephthalate complex, there are CHCl₃ molecules blocking the exits of the cylinder in the succinate complex. The CHCl₃ forms weak H-bonds with the carboxylate oxygen atoms (C⋯O = 3.167 and 3.222 Å). A similar circumstance was also observed in our group for FHF⁻ bound by a bicyclic cyclophane host.¹⁰ There are no water molecules in the side-pocket this time, rather there is a water slightly above the opening that forms a H-bond link to a neighboring host (see ESI†).

Fig. 4 ORTEP at 30% ellipsoid probability (top) and space-filling (bottom) views of the succinate complex with 1. Only amide H atoms for the host are shown in the ORTEP view on the top, and the Me₄N⁺ counterions and nonessential external solvent molecules are omitted in both views for clarity.

Solution characterization of the free ligand and complexed 1 were carried out using ¹H NMR spectroscopy. As anticipated the ¹H NMR spectrum of 1 displays two sets of amide signals in a 1 : 2 intensity ratio. These are indicative of the two different environments, that of the linker (four amides) and that of the two macrocycles (eight amides) (Fig. 5a). Furthermore, the NMR findings indicate that the dicarboxylates are also encapsulated in solution, as illustrated by the NMR spectra of three aromatic dicarboxylates, terephthalate, isophthalate and the extended 2,6-naphthalenedicarboxylate, as shown in Fig. 5. In the terephthalate case (Fig. 5b), the macrocyclic NH^a signal shows a significant downfield shift (Δδ = 1.67 ppm), while the amide spacers NH^a′ barely shift at all from the free base (Fig. 5a). In the isophthalate complex (Fig. 5d), there are actually two new signals assigned to macrocyclic NH^a protons at 10.55 and 10.25 ppm, indicating an unsymmetrical binding mode for the two macrocycles. In this case, the NH^a′ signal also shifts slightly downfield, possibly indicating some interaction of the guest with the spacer amides. The NH^a signal in 2,6-naphthalenedicarboxylate, also shows a significant downfield shift (Δδ = 1.82 ppm) (Fig. 5f); however, the other amide NH^a′ shifts upfield (Δδ = −0.31 ppm), possibly due to a shielding effect of the naphthalene rings.

Fig. 5 ¹H NMR spectra (400 MHz, 293 K, DMSO-d₆) of (a) 1, (b) 1·terephthalate, (c) terephthalate, (d) 1·isophthalate, (e) isophthalate, (f) 1·2,6-naphthalenedicarboxylate, (g) 2,6-naphthalenedicarboxylate.

The binding of 1 with the n-Bu₄N⁺ salts of selected dicarboxylates was evaluated by ¹H NMR titrations in DMSO-d₆.¹¹ With only one exception, all of the dicarboxylates tested, both aliphatic, from oxalate to suberate, and aromatic, showed slow exchange on the NMR timescale and gave K > 10⁵ M⁻¹ (see ESI†). The exception was 4,4′-biphenyldicarboxylate, for which K = 2.75 × 10⁴ M⁻¹. These findings, along with the crystallographic results, suggest that the host possesses an innate flexibility that allows for adjustment in length in order to mold to the length of a variety of dicarboxylate ions. The binding ability of 1 decreases sharply towards monocarboxylates such as acetate (K = 452 M⁻¹) and benzoate (K = 246 M⁻¹).

In conclusion, the shape-shifting amide-based tricyclic receptor, 1, displays an unusual aptitude for encapsulating dicarboxylates by molding itself to its guest. The host binds a broad variety of both aliphatic and aromatic dicarboxylate anions with extremely high affinities, most probably also by encapsulation, as indicated by the NMR data. The presence of side pockets for additional guests, as observed for the terephthalate complex, could lead to further applications as catalysts for chemical reactions by bringing reacting molecules into proximity. Finally, the extreme changes in the molecular dimensions for the p-carbon atoms of the pyridine units upon binding (>10 Å) are impressive and suggestive of further applications of similar hosts as molecular machines. The utility of this new tricyclic host for other multi-topic guests and the design of the other covalent organic molecular cages are currently underway.

The authors thank the National Science Foundation, CHE-0809736, for support of this work and CHE-0923449 for purchase of the X-ray diffractometer.

References

1. J.-M. Lehn, E. Sonveaux and A. K. Willard, J. Am. Chem. Soc., 1978, 100, 4914–4916; B. Dietrich, J. Guilhem, J.-M. Lehn, C. Pascard and E. Sonveaux, Helv. Chim. Acta, 1984, 67, 91–104; J.-M. Lehn, R. Méric, J.-P. Vigneron, I. Bkouche-Waksman and C. Pascard, J. Chem. Soc., Chem. Commun., 1991, 62–64.
2. M. W. Hosseini and J.-M. Lehn, J. Am. Chem. Soc., 1982, 104, 3525–3527; M. W. Hosseini and J.-M. Lehn, Helv. Chim. Acta, 1986, 69, 587–603.
3. R. J. Fitzmaurice, G. M. Kyne, D. Douheret and J. D. Kilburn, J. Chem. Soc., Perkin Trans. 1, 2002, 841–864 and references therein.
4. A. Bianchi, K. Bowman-James and E. García-España, Supramolecular Chemistry of Anions, Wiley-VCH, New York, 1997; I. Stibor, Anion Sensing. Topics in Current Chemistry, Springer, Berlin, 2005; J. L. Sessler, P. A. Gale and W.-S. Cho, Anion Receptor Chemistry, RSC Publishing, Cambridge, UK, 2006; Supramolecular Chemistry of Anionic Species, Themed Issue Chem. Soc. Rev., 2010, 3910.
5. M. M. Boyle, R. A. Smaldone, A. C. Whalley, M. W. Ambrogio, Y. Y. Botros and J. F. Stoddart, Chem. Sci., 2011, 2, 204–210; J.-P. Sauvage, J.-P. Collin, S. Durot, J. Frey, V. Heitz, A. Sour and C. Tock, C. R. Chim., 2010, 13, 315–328; N. N. P. Moonen, A. H. Flood, J. M. Fernández and J. F. Stoddart, Top. Curr. Chem., 2005, 262, 99–132; V. Balzani, M. Gómez-López and J. F. Stoddart, Acc. Chem. Res., 1998, 31, 405–414.
6. V. Maslak, Z. Yan, S. Xia, J. Gallucci, C. M. Hadad and J. D. Badjić, J. Am. Chem. Soc., 2006, 128, 5887–5894; B.-Y. Wang, S. Rieth and J. D. Badjić, J. Am. Chem. Soc., 2009, 131, 7250–7252.
7. M. C. Jimenez-Molero, C. Dietrich-Buchecker and J.-P. Sauvage, Chem.–Eur. J., 2002, 8, 1456–1466.
8. S. O. Kang, M. A. Hossain and K. Bowman-James, Coord. Chem. Rev., 2006, 250, 3038–3052.
9. S. O. Kang, D. Powell, V. W. Day and K. Bowman-James, Angew. Chem., Int. Ed., 2006, 45, 1921–1925; S. O. Kang, V. W. Day and K. Bowman-James, Inorg. Chem., 2010, 49, 8629–8636.
10. S. O. Kang, V. W. Day and K. Bowman-James, Org. Lett., 2008, 10, 2667–2680.
11. For fast exchange processes, the binding constants were calculated by EQNMR, see: M. Hynes, J. Chem. Soc., Dalton Trans., 1993, 311–312.

---

Footnote

† Electronic supplementary information (ESI) available: Crystallographic data in CIF format. Synthetic details, ¹H NMR titration spectra and crystallographic information. CCDC reference numbers 825550, 825551, 825552. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c1sc00292a

---