CationâAnion Arrangement Patterns in Self-Assembled Pd2L4 and

Article pubs.acs.org/accounts

Cation−Anion Arrangement Patterns in Self-Assembled Pd2L4 and Pd4L8 Coordination Cages Guido H. Clever* and Philip Punt Department of Chemistry and Chemical Biology, TU Dortmund University, Otto-Hahn-Str. 6, 44227 Dortmund, Germany CONSPECTUS: Compounds featuring one-dimensional regular arrangements of stacked metal complexes and alternating [cation−anion]∞ sequences have raised considerable interest owing to their peculiar electronic and optical properties as well as guest inclusion capabilities. While traditional ways to realize these structural motifs rely on crystalline compounds, exclusively existing in the solid state, recent progress in the area of metalmediated supramolecular self-assembly allows for the rational synthesis of structurally well-defined short stretches of stacked metal complexes and cation−anion arrangements. Therefore, metal cations, counteranions, and suitably designed organic bridges are allowed to self-assemble in solution. While the bridges can be designed as crosslinkers to yield extended two- or three-dimensional networks such as layered materials, metal−organic frameworks (MOFs), or porous coordination polymers (PCPs), they can also be tailored to lead to discrete nanoscopic objects. Supramolecular helicates, grids, and knots belong to this class of compounds, and a particularly interesting subfamily are coordination cages and capsules, which possess nanosized cavities with the ability to encapsulate guest molecules. Here, we focus on coordination cages consisting of two or more square-planar Pd(II) or Pt(II) metal cations, bridged by bananashaped bis-monodentate pyridyl ligands that encapsulate various guest molecules, usually anions, in their cavities. Monoanions as well as dianions with localized or delocalized charges can be bound with remarkable complementarity between cage and guest in terms of size and shape. We show how dimerization of the prototypical [Pd2L4] cages into their interpenetrated dimers [Pd4L8] leads to an increase in cavity number from one to three while the cavity volume decreases. Usually, all three pockets of these double cages are filled with monoanions such as BF4− or Cl−, thus leading to well-defined linear [Pd−anion]3Pd stacks, as observed by X-ray studies. The ligand-based mechanical coupling of the linearly aligned cavities leads to interesting effects concerning guest encapsulation cooperativity, such as allosteric binding and triggered sequential uptake. While most of the so far reported coordination cages consist of only a single type of ligand, recent advances in rational assembly strategies allow for high-yielding syntheses of structurally defined multicomponent architectures by integrative self-sorting mechanisms. One family of heteroleptic [Pd2L2L′2] cages whose formation is based on shape-complementarity between two different ligands, L and L′, is introduced. Furthermore, the implementation of ligand-based functions such as redox activity, photochromic behavior, specific binding sites, chirality, and catalytic activity allows us to study systems with properties far beyond basic structural features. We showcase selected examples of self-assembled cages whose guest uptake or even overall structural integrity is reversibly switched by light or small molecules with potential application in stimuli responsive materials (e.g., for sequestration of pollutants or stabilization of reactive compounds) up to functional nanosystems (e.g., diagnostic devices or supramolecular catalysts) and molecular machines.

1. INTRODUCTION Columnar arrangements of metal complexes and alternating cation−anion stacks are among the most interesting structural motifs found in solid state structures of inorganic compounds, since metal−metal interactions or anion-bridged metal centers often lead to valuable material properties.1−3 In addition, extended coordination compounds such as metal−organic frameworks allow for the uptake of guest molecules in their porous structure.4−6 Owing to their solid state nature, however, this property is restricted to bulk materials, deposited surfaces, and suspensions. This Account, on the other hand, puts a focus on soluble coordination architectures of defined stoichiometry and size. In particular, sequences of alternating cation−anion arrangements in the form of supramolecular self-assemblies are discussed, starting from the shortest stacks consisting of © 2017 American Chemical Society

only two cations sandwiching one anion to systems with up to six metal centers. For discrete assemblies of one-dimensional metal wires with direct metal−metal contacts, the reader is referred to the work of Berry, Cotton,7 and Peng.8 Metal−DNA architectures9,10 and peptide-stabilized11 metal stacks are also not covered here. Rather, this Account will introduce a family of artificial coordination cages12−14 in which oppositely arranged square-planar metal cations are connected by banana-shaped bis-monodentate ligands to give hollow [M2L4] cages that, in the Pd(II)- and Pt(II)-based systems discussed here,15−21 possess an overall tetracationic charge when the ligands are neutral molecules such as pyridines. Consequently, an internal cavity is Received: May 7, 2017 Published: August 17, 2017 2233

DOI: 10.1021/acs.accounts.7b00231 Acc. Chem. Res. 2017, 50, 2233−2243

Article

Accounts of Chemical Research

Figure 1. Schematic representation of anion-containing cationic Pd(II)based coordination cages of different shape and nuclearity: (a) prototypical [Pd2L4] motif including one bis-anionic guest; (b) the same cage encapsulating a pair of anions sandwiching a cationic guest; (c) interpenetrated [Pd4L8] motif with three anion-filled cavities; (d) two anions inside a flexible [Pd2L4] cage; (e) bent bis-anion inside a bent heteroleptic [Pd2L2L′2] cage; (f) stack of six Pd(II) cations in the solid state structure of a triple catenated metallo-architecture.

Figure 2. (a) Self-assembly of banana-shaped ligands with square-planar Pd(II) cations to monomeric [Pd2L4] cages (capable of binding larger or bis-anionic guests) and further dimerization into interpenetrated double cages [Pd4L8] (binding three anionic guests in their three pockets); (b) trinuclear and (c) tetranuclear ring structures; (d) unique doubletrefoil knot structure.

formed that is able to encapsulate guest molecules, and usually anionic guests are bound with high affinity. Figure 1a schematically sketches the encapsulation of a bis-anionic species in which the negatively charged ends (yellow spheres) are attached to central parts of different lengths that allow for probing the relationship between host and guest size. In contrast, also two individual anionic molecules can be encapsulated and a special case is shown in Figure 1b where an additional cationic guest separates the two anionic species inside the cavity, hence leading to a consecutive cation−anion−cation−anion−cation stack. On the other hand, stacks of multiple cationic and anionic units were also found to form in interpenetrated [Pd4L8] dimers of the aforementioned Pd(II) (but not Pt(II)!) cages (Figure 1c).16 Flexible, bent, and catenated systems depicted in Figure 1d−f are discussed in detail below.

2. CAGE ASSEMBLY, GUEST UPTAKE, AND INTERNAL DYNAMICS 2.1. Cage Formation

To synthesize the cages, ligands, and a palladium salt with noncoordinating anions (e.g., [Pd(CH3CN)4](BF4)2) are mixed in a polar solvent such as acetonitrile or DMSO. Depending on the ligand design, a single self-assembled product or a mixture of compounds of different nuclearity is formed (e.g., 3- and 4-rings; Figure 2).15 Entangled structures such as interpenetrated dimers (Figure 2a, right)16 or knots (Figure 2d)22 take considerably longer to form since they are preceded by the corresponding smallest possible [Pd2L4] cages as kinetic products, which subsequently combine to the interwoven architectures. Questions concerning the driving forces responsible for the formation of specific ring sizes or entangled structures, as depending on the ligand structure, chosen solvent, and counteranions, can only be answered in parts. First of all, entropic reasons favor the lowest nuclear

Figure 3. Early examples of [Pd2L4] coordination cages: (a) PF6− encapsulated in a cavity based on ligands with a phenyl ether backbone. (b) Chloride-containing cage with flexible amide ligands.

assembly in solution since this leads to formation of the highest number of independent particles. This is also the reason for the observation that the formation of polymers is disfavored. Furthermore, entropic factors play an important role in connection with solvent effects; in particular the release of cavity-confined solvent molecules favors dimerization. In terms of enthalpic contributions, cage formation is mainly driven by replacing Pd-coordinated solvent molecules by 2234


Article

Accounts of Chemical Research pyridine donors. Concerning cage dimerization, the formation of one-dimensional alternating [cation−anion]3−cation stacks in the interpenetrated [Pd4L8] double cages seems to be a favorable arrangement. It has to be mentioned, however, that four dications are combined with three monoanions, leaving the “filled” double cages with a formal charge of +5. It was further shown that BF4− anions in the outer pockets rapidly exchange with the anions in free solution. Finally, in the case of the acridone-based cage [3BF4@Pd4L128] (see below), even the central BF4− anion can get lost when the outer pockets bind halide anions. As X-ray structures show, chloride anions encapsulated in the double cages are surrounded by eight hydrogen atoms in a square antiprismatic geometry with Harom−Cl− distances between 2.5 and 2.7 Å.16 Further enthalpic stabilization may be contributed by π-stacking between the ligand arms of the entangled cage subcomponents. Finally, we could also observe pronounced kinetic stabilization effects of the interpenetrated double cages toward ligand

exchange in mixing experiments: While isostructural monomeric [Pd2L4] cages quickly exchange ligands in mixing experiments, once-formed interpenetrated double cages resist exchanging ligands with other double cages over days and weeks.23 2.2. Host−Guest Chemistry in Homoleptic Monomeric [Pd2L4] Cages

The monomeric [Pd2L4] cages possess one cavity whose size is dependent on the dimensions of the ligand. The earliest reported example by McMorran and Steel (Figure 3a), based on 1,4-bis (3-pyridyloxy)benzene ligand L1, turned out to possess the perfect dimensions to encapsulate a single hexafluorophosphate anion.24 Other early work in guest encapsulation was contributed by Puddephatt using bis(amidopyridine) ligand L2 to encapsulate a chloride anion (Figure 3b).19 Later examples increased constantly in size. Most of the reported coordination cages based on the Pd(pyridine)4 or cis-Pd(ethylenediamine)(pyridine)2 cis-protected motifs, many of which were contributed by the groups around Fujita12 and Stang,13 made use of primarily bent

Figure 4. Assembly and guest encapsulation of selected monomeric cages [M2L4] (M = Pd, Pt). (a) First-generation oligonorbornane cage and its encapsulation of a pair of square-planar bis-anions together with a flat cation to give discrete one-dimensional Pt5-stacks. (b) Desymmetrization of an octahedral PtIV-complex inside a hydrogen-bonded donor-carrying oligonorbornane cage. (c) Flipping dynamics and guest encapsulation inside a sterically overcrowded adamantyl cage and (d) rotational dynamics in interplay with guest binding in a push−pull cage system. 2235


Article


bulky substituents in endohedral positions. Initially, these studies aimed at preventing cage dimerization in order to access cages with increased cavity sizes. Interestingly, substituents connected by a double bond to the inner position of the backbone were shown to exhibit pronounced dynamic effects. First, we investigated ligand L5 carrying a bulky adamantyl group in the endohedral position (Figure 4c).33 Steric strain pushes the bulky group toward one ligand face but still allows for flipping with a moderate rate (150 s−1) between the two degenerate conformations of the ligand. This ligand cleanly forms [Pd2L54] cages when mixed with Pd(II) cations. X-ray structure analysis reveals that the adamantyl groups adapt positions in the cage’s portals by obeying a common rotational sense. They are still able to flip within the cage architecture, albeit at a largely diminished rate (0.32 s−1), and this seems to happen without breaking the metal−ligand bonds as the corresponding platinum cage behaves similarly to its palladium relative. Since molecular modeling studies prohibit a concerted flipping, we assume a stepwise flipping mechanism. To our great surprise, even this sterically congested cage is able to encapsulate guest molecules of quite substantial dimensions. Figure 4c shows the encapsulation of the bulky 1,1′ferrocene bis-sulfonate as proven by NMR titration experiments, ESI mass spectrometry, and via X-ray diffraction analysis. One member of a slightly less sterically demanding class of substituents is shown in Figure 4d. Ligand L6 consists of a malononitrile ester moiety of which we studied several derivatives with respect to the type of ester.34 Whereas this substituent still shows the aforementioned flipping, it also exhibits rotation around the formal double bond. The latter effect is based on its electronic push−pull character, consisting of the electron-donating amine in the backbone and the electronwithdrawing nitrile and ester substituents at the double bond. Interestingly, guest uptake was found to slow down the spinning of the endohedral rotors.

oligo-arylene ligands. In contrast, we investigated in our early work the complexation behavior of aliphatic ligands such as L3 whose oligonorbornene-like backbone gives the required curvature and provides substantial bulkiness to prevent dimerization of the cages (Figure 4a).25 A novelty of our ligand design was its ribbon shape, which not only affected cage stiffness and guest binding but also allowed for the unambiguous NMRbased determination of the knot structure shown in Figure 2d.22 The resulting cages [Pd2L34] are rigid and carry the two palladium cations in a distance of about 17 Å acting as cationic “anchors” for the binding of anionic guests. Several types of anions were probed for encapsulation, among them aromatic bis-sulfonates, polyoxometallates, and charged square-planar Pt-complexes. For a series of aromatic bis-sulfonates we could show that the binding affinity rises with the distance between the anionic groups because of increased Coulomb attraction with the metal centers.26 There is, however, a length limit dictated by the cage dimensions that cannot be exceeded. The latter observation allowed us to control uptake and release of a light-switchable guest that changes its size.27 The successful encapsulation of a hexamolybdate bis-anion, Mo6O192−, inside cage [Pd2L34]4+ demonstrated that the guest scope can be extended to inorganic, metal-based compounds.28 Figure 4a shows another encapsulation experiment with metal-based guests, this time in the platinum-variant, [Pt2L34]2+, of the cage. Here, two square-planar tetrachloroplatinate bis-anions are coencapsulated together with a [Pt(pyridine)4]2+ cation to yield an alternating, one-dimensional arrangement of cationic and anionic [PtIIligand4] units that mimics a short sequence of the colored Magnus’s salts, which consist of infinite stacks of alternating cationic and anionic square-planar Pt(II) complexes.29 Confinement into the boundaries of the cage thus allows for the preparation of isolated Pt5-stacks in the form of soluble supramolecules, a strategy that may facilitate the processing and application of such electronically interesting stacked compounds.30 In collaboration with the Pfeffer group, we obtained a series of new ligand derivatives31 of which one member, L4, is depicted in Figure 4b.32 Apart from its adventurous structure, its distinctive feature is the endohedral functionality pointing right into the middle of the cavity. Here, this functionality is a hydrogen-bond donor site, but a range of other functional group attachments is accessible. Besides its two cationic Pd(II) anchors occupying the axial positions, cage [Pd2L44]4+ features four hydrogen-bond donors arranged in a square on the axial plane dissecting the cage into two halves. This peculiar arrangement of binding sites in 3-dimensional space prompted us to investigate the encapsulation of linear, square-planar, and octahedral metal-based guests. Interestingly, linear ([Ag(CN)2]−) and square-planar guests ([PtII(CN)4]2−) were bound more weakly than octahedral ones ([PtIV(CN)6]2−, [Fe(CN)6]3−), and bis-anionic guests were bound more strongly than monoanionic (PF6−) or neutral guests ([M(CO)6]0; M = Cr, Mo, W). Most interestingly, however, was the observation that the tetragonal environment of cage [Pd2L44], acting as a second coordination sphere around encapsulated hexacyanoplatinate(IV), can imprint its D4h-symmetry on the originally Oh-symmetric metal complex and accordingly reduce its symmetry in the host−guest complex, unambiguously supported by X-ray and IR-spectroscopic examinations.

3. INCREASING FUNCTIONAL COMPLEXITY Next, we set out to surpass the passive functionality showcased in the previous section by introducing actively controllable

Figure 5. Dithienylethene-based photochromic cage [Pd2L74] can be switched between a conformationally flexible and a rigid photoisomer, which show different binding affinities for globular anionic guests. (a) Irradiation with light of different wavelengths allows reversible switching between the high-affinity open-form photoisomer and the low-affinity closed-form. (b) Chemical structure of both photoisomeric forms indicating the conformational flexibility of o-L7.

2.3. Endohedral Functionality and Dynamics

Another class of homoleptic monomeric [Pd2L4] cages was constructed from a flat aromatic backbone (that otherwise would form interpenetrated double cages as shown below) by attaching 2236


Article


wavelengths (Figure 5). Both cages were found to bind the globular [B12F12]2− bis-anion, but the open-form photoisomer (whose X-ray structure with BF4− counteranions shows a steplike shape) was shown to bind the guest about 100 times more strongly than the closed-form photoisomer. Irradiation in the presence of the guests then allowed for shifting the guest encapsulation equilibrium more toward the host−guest complex or the free cage, respectively. Ligand L8 differs from L7 in only two points: first, it is shorter since no alkyne linker is present, and second its pyridines are attached to the backbone with a para-orientation (Figure 6a).36 The open form of ligand L8 was found to form the 3-ring structure [Pd3o-L86] (in equilibrium with some 4-ring [Pd4o-L88]). In these rings, the flexible DTE backbone is highly twisted, thus deviating from the usual lowest energy conformations of the free ligands. The closed-form ligand, however, is a rigid molecule with

functions in the backbone structure in order to obtain stimuliresponsive systems. Light was chosen as the preferred external trigger as it can be dosed with temporal and spatial control and leaves no traces. 3.1. Light-Switchable Coordination Cages

The construction of a light-responsive supramolecular system requires the implementation of a robust and high-performing photoswitch, and we chose the diarylethene derivative L7 for its fatigue-resistance, cleanly accessible photostationary states, and synthetic suitability to append coordinating donor functionalities.35 Ligand L7 could be prepared in the form of the open and closed isomer (with respect to the central ring) where the open form shows a pronounced conformational flexibility in its backbone while the closed form has a rigid backbone structure. Pleasingly, both photoisomers readily form [Pd2L74] cages, which can be interconverted by irradiation with light of different

Figure 6. (a) Photoswitching and structural conversion of small [Pd3o-L86] and [Pd4o-L88] rings into large [Pd24c-L848] spheres and vice versa. (b) DOSY NMR comparison of rings and spheres. (c) TEM visualization of the uniformly sized sphere particles and (d) comparison of photoswitch conformations unfavorable (top, as in the small rings) and favorable (bottom, as in the sphere) for photoswitching. Reproduced with permission from ref 36. Copyright 2016 Wiley-VCH. 2237


Article

Accounts of Chemical Research a wide angle between the pyridine donor sites. Consequently, coordination to Pd(II) cations must lead to an assembly with a larger curvature and, as shown by the systematic investigations of Makoto Fujita,37 obey a stoichiometry [PdnL2n] with n taking only certain numbers for geometric reasons. We showed for the first time, that light-switchable ligands can form the basis of such a large polyhedral assembly. The formation of a large [Pd24c-L848] sphere, about 7 nm in diameter, was indicated by DOSY (diffusion ordered spectroscopy) NMR spectroscopy (Figure 6b), cryospray ESI mass spectrometry, atomic force microscopy, and small-angle X-ray scattering. Transmission electron microscopy (TEM, Figure 6c) revealed uniform particles of about 7 nm size as expected. An interesting observation was made when comparing the switching kinetics of the assemblies based on ligands L7 and L8. In the L7 system, switching is comparably fast for the free ligands and the assembled cages. In the case of the L8-based architectures, the photochemical switching rate for going from the large sphere containing only the closed-form photoisomers of the ligand is comparable to free ligand L8 (as well as to the L7 system). The reverse process, however, light-triggered conversion of the small rings into the large sphere, is slower by several orders of magnitude. The reason is found in the peculiar conformational locking of the open-form photoswitches in the small rings, which distorts the conjugated π-system in such a way that photoexcitation cannot result in formation of the closed-form photoisomer. Therefore, photoisomerization does not proceed from the ring-shaped assemblies but from the small steady-state concentration of free ligand that is always present in the dynamic equilibrium mixture. 3.2. Allosteric and Sequential Guest Binding in Interpenetrated Double Cages

Less bulky banana-shaped ligands of sufficient length show a strong tendency to form interpenetrated double cages. While Sekija and Kuroda had previously reported the formation of a first example of a [Pd4L98] double cage based on a conformationally flexible bis-monodentate ligand L9 (Figure 7a),38 this dimer was thermodynamically only slightly favored over the monomer in the corresponding equilibrium. Our investigations in this area started off with a more rigid backbone, commercially available dibenzosuberone, which can be easily equipped with the needed pyridine arms on one side of the tricyclic core. Whereas a short version of the ligand with direct 3-pyridyl attachments forms monomeric [Pd2L4] cages, the longer variant L10 forms a monomeric cage as a kinetic intermediate (Figure 7b), which then quantitatively converts into the interpenetrated dimers carrying three anions sandwiched between the four metal centers.39 This reaction, usually carried out in a polar solvent such as acetonitrile, is based on [Pd(CH3CN)4](BF4)2 as a source for palladium(II). Tetrafluoroborate was chosen as counteranion due to its noncoordinative character and appropriate size for filling the pockets of the double cage. After some hours of heating the reaction mixture, the formation of the interpenetrated double cages [3BF4@Pd4L108]5+ proceeded virtually quantitatively as monitored by 1H NMR spectroscopy, ESI mass spectrometry, and X-ray structure determination. Interestingly, this system was found to be a strong receptor for halide anions, which bind with positive cooperativity in the outer two pockets via an allosteric binding mechanism that is accompanied by a compression of the double cage along the Pd4-axis.40 As a result, the outer pockets in [2Cl+BF4@ Pd4L108]5+ shrink to accommodate the halide anions sandwiched

Figure 7. (a) The first reported interpenetrated double cage of the type [Pd4L98] by Kuroda et al. (b, c) Assembly and allosteric guest binding in two related interpenetrated double cages with preference for (b) small halides and (c) larger anions controlled by the size of the central template.

between the Pd(pyridine)4-planes in the optimal distance and the inner pocket expands as confirmed by a change in 19F relaxation kinetics of the encapsulated BF4− anion that senses the change in available cavity space.40 We then looked into tuning the anion selectivity (i.e., chloride vs bromide) by systematically 2238


Article


was rebutted by the observation that also cyclohexane is bound, though slower, but even with a higher affinity than benzene (Figure 8b). Besides chloride, also bromide binding can trigger the uptake of the neutral guests, but with altered uptake kinetics and thermodynamics. Triggered binding of small molecules in these systems opens perspectives for application in supramolecular sensing, signal transduction, and catalysis under external control. The same triggers, small halide anions, were found to induce an entirely different process in the following system: ligand L13,

changing the ligand length but found that chloride always binds more strongly than bromide, which could be subsequently explained by computational results.41 Interestingly, chloride binding is so strong that the cages can dissolve solid AgCl in acetonitrile (leading to an estimated net binding affinity of ∼1020 M−2).40 A further modification of the ligand backbone was found to yield a double cage system that indeed has a very different anion preference. Therefore, ligand L11 was synthesized by adding an aromatic residue to the carbonyl group of precursor L10 thereby introducing steric bulk that does initially not allow the formation of an interpenetrated double cage upon reaction with palladium tetrafluoroborate (Figure 7c).42 To our delight, the addition of chloride as a small anionic template subsequently allowed dimerization since the small halide anion (now sitting in the central pocket!) permitted the two cage subunits in [Cl@Pd4L118]7+ to keep a large enough distance that the bulky aryl substituents would not clash. Consequently, large outer pockets are formed that allow for the selective and allosteric encapsulation of large anions such as perrhenate. In comparison both systems can be regarded as an adjustable anion binding system in which the size of the central template is reversely proportional to the size of the preferred guests in the outer pockets. 3.3. Stimuli-Responsive Double Cages

Initially, the examination of a further ligand derivative L12 based on the heterocycle acridone followed the same pattern as discussed above for the dibenzosuberone system with formation of a double cage and allosteric halide binding in the outer two pockets (Figure 8a).43 We solved molecular structures via X-ray

Figure 9. (a) Stepwise, halide-triggered structural reorganization of a carbazole-based single cage into its interpenetrated dimer and a neutral triple catenane and (b−d) molecular structures.

Figure 8. (a) Assembly and chloride-triggered uptake of small neutral guests of an acridone-based double cage, (b) comparison of uptake kinetics and thermodynamics, and (c, d) X-ray structures of [3BF4@ Pd4L128]5+ and [2Cl+benzene@Pd4L128]6+.

diffraction analysis of both the as-synthesized compound [3BF4@Pd4L128]5+ and its chloride-bound successor where in both cases the crystals were grown from vapor diffusion of benzene into an acetonitrile solution of the cages. To our surprise, the latter compound now contained a single benzene molecule in its central pocket, while both structures showed additional cocrystallized benzenes outside the cage perimeters. Species [2Cl+benzene@Pd4L128]6+, in which a neutral molecule has replaced an anion from the pocket of a highly cationic host, was also unambiguously identified by NMR spectroscopy and mass spectrometry. Our initial assumption that cation−π interactions were responsible for uptake of the aromatic guest

Figure 10. (a) The combination of isostructural electron-rich phenothiazine donor ligand L14 and electron-poor anthraquinone acceptor ligand L15 can be realized in the form of a mixture of two homoleptic double cages or as a statistical mixture of mixed double cages containing all possible stoichiometries and stereoisomers. (b) Photoexcitation of the mixed double cages leads to light-induced charge separation as evidenced by (c) time-resolved pump−probe spectrosco2239


Article


4. MULTIFUNCTIONAL HETEROLEPTIC ASSEMBLIES All self-assembled compounds introduced above consisted of one-type of ligand, each. Next, we became interested in mixing at least two functionalities within one single cage assembly in order to study cooperative and emergent behavior that arises from the combination of the different ligands.

based on carbazole as a backbone, is slightly too short to allow the formation of a double cage templated by the available BF4− anions (Figure 9).44 We were, however, able to induce dimerization by offering rather small halide anions (X = Br− or Cl−) as templates that now occupy all three pockets of double cage [3X@Pd4L138]5+. Interestingly, increasing the concentration of the same halide trigger was found to lead to yet another structural interconversion by replacing two pyridine ligands on each palladium center. Thereby, cages transform into trans[(PdX2)L132] rings, and while we cannot exclude that such individual rings exist in solution, the solid-state structure for X = Br− delivered a rather spectacular triple catenated architecture {trans-[(PdBr2)L132]}3 in which all six neutral PdBr2 units were found to stack in a quasi-one-dimensional fashion.

4.1. Mixed Cages for Light-Induced Charge Transfer

Our first endeavor in the area of mixed cages was inspired by previous work by us45 and others46 demonstrating that coordination cages can be equipped with redox-active moieties and electrochemically converted into oxidized or reduced species. We then turned our focus on mixtures of isostructural ligands with different electron demand (Figure 10).47 Whereas

Figure 11. Systematic assembly studies with bis-monodentate ligands that deviate from the ideal banana-shape. (a) Inward-bent ligand L16 and (b) outward-pointing ligand L17 form homoleptic [Pd2L164] cages and [Pd4L178] rings, respectively. (c) When mixed 1:1, they cleanly form heteroleptic bent cage [Pd2L162L172]. (d) A similar bent cage [Pd2L132L172] is formed from ligands L13 and L17, whereas (e) L13 combined with L16 leads to the unprecedented “doubly-bridged figure-eight” compound trans-[Pd2L132(anti-L16)2]. 2240



■

both the electron donor ligand L14, based on phenothiazine, and acceptor ligand L15, based on anthraquinone, are able to form stable homoleptic double cages that do not exchange ligands when mixed, the exposure of a ligand mixture to Pd(II) cations leads to a statistical mixture of double cages with all possible ligand ratios and stereochemical relationships. Interestingly, light excitation of the donor in these mixed cages leads to a charge separated state for which femtosecond time-resolved UV−vis spectroscopy revealed a signature that matches the sum of the spectra of the oxidized donor and reduced acceptor as obtained by a spectro-electrochemical characterization of the homoleptic cages (Figure 10c). Such a close packing of functional units in a mechanically locked architecture might inspire new structural motifs for the rational design of photovoltaic bulk-heterojunctions with controlled donor−acceptor morphology.

Article

AUTHOR INFORMATION

Corresponding Author

* E-mail: [email protected]. Website: www.clever-lab.de ORCID

Guido H. Clever: 0000-0001-8458-3060 Funding

The DFG (489/2-1, SFB 1073, SPP 1807), the CaSuS Program of Lower Saxony, and the Fonds der Chemischen Industrie are greatly acknowledged for their support. Notes

The authors declare no competing financial interest. Biographies Guido H. Clever, born in 1976, studied chemistry in Heidelberg, changed to Marburg University, and received his Ph.D. in bioorganic chemistry from LMU Munich in 2006. From 2007−2009, he was an AvH/JSPS postdoctoral researcher and from 2009 an Assistant Professor at The University of Tokyo. In 2010, he joined the University of Göttingen as a Junior-Professor where he was appointed as a Professor in 2013. In 2015, he became a full professor at TU Dortmund. His research focuses on the redox-, photo-, and host−guest chemistry of functionalized coordination cages. Further interests include metalmediated DNA and chiral organometallic chromophores.

4.2. Rational Assembly Strategies

In order to overcome the shortcomings of statistical self-assembly governed by entropy, more sophisticated assembly strategies are currently being worked out by us and others.48−51 Examples are based on charge balancing, additional hydrogen bonding, macrocyclic preorganization, and donor site engineering. In our approach, pairs of shape complementary bis-pyridyl ligands are combined with Pd(II) cations in a 1:1:1 ratio to cleanly assemble heteroleptic cages (Figure 11). For example, ligands L16 and L17, both being able to form homoleptic [Pd2L164] cages and [Pd4L178] rings, respectively (Figure 11a,b), together form bent cage [Pd2L162L172] in which the two differently shaped ligands are arranged in a cis-fashion (Figure 11c).52 Owing to its bent shape, this cage prefers binding bent guests (see Figure 1e). Likewise, ligands L13 and L17 were found to cleanly assemble heteroleptic cage [Pd2L132L172] (Figure 11d). Surprisingly, the combination of the allegedly non-shape-complementary ligands L13 and L16 yielded an unprecedented trans-[Pd2L132(anti-L16)2] assembly whose geometry can be described as a “doubly-bridged figureeight” (Figure 11e).53 For further discussion of multicomponent design strategies, the reader is referred to a recent feature article about this topic.54

Philip Punt was born in 1992. He studied Chemical Biology at TU Dortmund University and obtained his bachelor degree in the area of cellular biology. He joined the Clever Lab in 2017 as a master student, which will soon be followed by a Ph.D. thesis on metalated DNA nanostructures.

■

ACKNOWLEDGMENTS I cordially thank my co-workers and collaborators who contributed to our work in the field of supramolecular selfassembly in the past years.

■

REFERENCES

(1) Chen, C.-T.; Suslick, K. S. One-Dimensional Coordination Polymers: Applications to Material Science. Coord. Chem. Rev. 1993, 128, 293−322. (2) Slabbert, C.; Rademeyer, M. One-Dimensional Halide-Bridged Polymers of Metal Cations with Mono-Heterocyclic Donor Ligands or Cations: a Review Correlating Chemical Composition, Connectivity and Chain Conformation. Coord. Chem. Rev. 2015, 288, 18−49. (3) Leong, W. L.; Vittal, J. J. One-Dimensional Coordination Polymers: Complexity and Diversity in Structures, Properties, and Applications. Chem. Rev. 2011, 111, 688−764. (4) Kitagawa, S.; Kitaura, R.; Noro, S.-I. Functional Porous Coordination Polymers. Angew. Chem., Int. Ed. 2004, 43, 2334−2375. (5) Zhou, H.-C.; Long, J. R.; Yaghi, O. M. Introduction to Metal− Organic Frameworks. Chem. Rev. 2012, 112, 673−674. (6) Schneemann, A.; Bon, V.; Schwedler, I.; Senkovska, I.; Kaskel, S.; Fischer, R. A. Flexible Metal-Organic Frameworks. Chem. Soc. Rev. 2014, 43, 6062−6096. (7) Berry, J. F.; Cotton, F. A.; Daniels, L. M.; Murillo, C. A. A Trinickel Dipyridylamido Complex with Metal−Metal Bonding Interaction: Prelude to Polynickel Molecular Wires and Devices? J. Am. Chem. Soc. 2002, 124, 3212−3213. (8) Ismayilov, R. H.; Wang, W. Z.; Lee, G.-H.; Yeh, C. Y.; Hua, S.-A.; Song, Y.; Rohmer, M. M.; Bénard, M.; Peng, S.-M. Two Linear Undecanickel Mixed-Valence Complexes: Increasing the Size and the Scope of the Electronic Properties of Nickel Metal Strings. Angew. Chem., Int. Ed. 2011, 50, 2045−2048. (9) Clever, G. H.; Kaul, C.; Carell, T. DNA-Metal Base Pairs. Angew. Chem., Int. Ed. 2007, 46, 6226−6236.

5. SUMMARY AND PROSPECTUS The self-assembly of banana-shaped bis-monodentate ligands with square-planar metal cations delivers a structurally and functionally diverse class of nanosized supramolecular structures. While linear arrangements of metal centers, commonly in alternation with interstitial anions, predominate, for example, in the interpenetrated double cages, also bent cages, knots, and tetrahedra,55 as well as rings and spheres of different size, can be obtained. Many of these systems show an interesting host−guest chemistry, from simple binding of single anionic guests over the allosteric formation of one-dimensional mixed-anion combinations to actively triggered guest uptake and release processes. Furthermore, research efforts in recent years focused on the implementation of specific function into the ligand backbones, thereby equipping these architectures with additional features such as photo- and redox-chemistry and catalytic activity. In parallel, strategies for rational, heteroleptic assembly are developed that allow for designing nanostructures of ever increasing complexity from scratch. Inspired by nature’s ability to produce dynamic and processive tools of breath-taking complexity, the combination of both of the latter topics will allow synthetic chemists to make further steps in creating multifunctional supramolecular architectures that act as enzyme mimics, nanoscale electronic, magnetic, and optical devices, and new smart materials. 2241


Article

Accounts of Chemical Research (10) Scharf, P.; Müller, J. Nucleic Acids with Metal-Mediated Base Pairs and Their Applications. ChemPlusChem 2013, 78, 20−34. (11) Tanaka, K.; Kaneko, K.; Watanabe, Y.; Shionoya, M. A TemplateDirected Synthetic Approach to Halogen-Bridged Mixed-Valence Platinum Complexes on Artificial Peptides in Solution. Dalton Trans. 2007, 5369−5371. (12) Fujita, M.; Tominaga, M.; Hori, A.; Therrien, B. Coordination Assemblies From a Pd(II)-Cornered Square Complex. Acc. Chem. Res. 2005, 38, 369−378. (13) Chakrabarty, R.; Mukherjee, P. S.; Stang, P. J. Supramolecular Coordination: Self-Assembly of Finite Two- and Three-Dimensional Ensembles. Chem. Rev. 2011, 111, 6810−6918. (14) Smulders, M. M. J.; Riddell, I. A.; Browne, C.; Nitschke, J. R. Building on Architectural Principles for Three-Dimensional Metallosupramolecular Construction. Chem. Soc. Rev. 2013, 42, 1728−1754. (15) Han, M.; Engelhard, D. M.; Clever, G. H. Self-Assembled Coordination Cages Based on Banana-Shaped Ligands. Chem. Soc. Rev. 2014, 43, 1848−1860. (16) Frank, M.; Johnstone, M. D.; Clever, G. H. Interpenetrated Cage Structures. Chem. - Eur. J. 2016, 22, 14104−14125. (17) Custelcean, R. Anion Encapsulation and Dynamics in SelfAssembled Coordination Cages. Chem. Soc. Rev. 2014, 43, 1813−1824. (18) Yoshizawa, M.; Yamashina, M. Coordination-Driven Nanostructures with Polyaromatic Shells. Chem. Lett. 2017, 46, 163−171. (19) Yue, N.; Eisler, D.; Jennings, M.; Puddephatt, R. A Lantern Complex of Palladium(II) as a Host for Anion Inclusion. Inorg. Chem. Commun. 2005, 8, 31−33. (20) Liao, P.; Langloss, B. W.; Johnson, A. M.; Knudsen, E. R.; Tham, F. S.; Julian, R. R.; Hooley, R. J. Two-Component Control of Guest Binding in a Self-Assembled Cage Molecule. Chem. Commun. 2010, 46, 4932−4934. (21) Sahoo, H. S.; Chand, D. K. Conformation of N,N′-Bis(3Pyridylformyl)Piperazine and Spontaneous Formation of a Saturated Quadruple Stranded Metallohelicate. Dalton Trans. 2010, 39, 7223− 7225. (22) Engelhard, D. M.; Freye, S.; Grohe, K.; John, M.; Clever, G. H. NMR-Based Structure Determination of an Intertwined Coordination Cage Resembling a Double Trefoil Knot. Angew. Chem., Int. Ed. 2012, 51, 4747−4750. (23) Frank, M.; Krause, L.; Herbst Irmer, R.; Stalke, D.; Clever, G. H. Narcissistic Self-Sorting vs. Statistic Ligand Shuffling Within a Series of Phenothiazine-Based Coordination Cages. Dalton Trans. 2014, 43, 4587−4592. (24) McMorran, D.; Steel, P. The First Coordinatively Saturated, Quadruply Stranded Helicate and Its Encapsulation of a Hexafluorophosphate Anion. Angew. Chem., Int. Ed. 1998, 37, 3295−3297. (25) Clever, G. H.; Tashiro, S.; Shionoya, M. Inclusion of Anionic Guests Inside a Molecular Cage with Palladium(II) Centers as Electrostatic Anchors. Angew. Chem., Int. Ed. 2009, 48, 7010−7012. (26) Clever, G. H.; Kawamura, W.; Shionoya, M. Encapsulation Versus Aggregation of Metal-Organic Cages Controlled by Guest Size Variation. Inorg. Chem. 2011, 50, 4689−4691. (27) Clever, G. H.; Tashiro, S.; Shionoya, M. Light-Triggered Crystallization of a Molecular Host-Guest Complex. J. Am. Chem. Soc. 2010, 132, 9973−9975. (28) Han, M.; Hey, J.; Kawamura, W.; Stalke, D.; Shionoya, M.; Clever, G. H. An Inclusion Complex of Hexamolybdate Inside a Supramolecular Cage and Its Structural Conversion. Inorg. Chem. 2012, 51, 9574−9576. (29) Clever, G. H.; Kawamura, W.; Tashiro, S.; Shiro, M.; Shionoya, M. Stacked Platinum Complexes of the Magnus’ Salt Type Inside a Coordination Cage. Angew. Chem., Int. Ed. 2012, 51, 2606−2609. (30) Klosterman, J. K.; Yamauchi, Y.; Fujita, M. Engineering Discrete Stacks of Aromatic Molecules. Chem. Soc. Rev. 2009, 38, 1714−1725. (31) Johnstone, M. D.; Schwarze, E. K.; Clever, G. H.; Pfeffer, F. M. Modular Synthesis of Linear Bis- and Tris-Monodentate Fused [6]Polynorbornane-Based Ligands and Their Assembly Into Coordination Cages. Chem. - Eur. J. 2015, 21, 3948−3955. (32) Johnstone, M. D.; Schwarze, E. K.; Ahrens, J.; Schwarzer, D.; Holstein, J. J.; Dittrich, B.; Pfeffer, F. M.; Clever, G. H.

Desymmetrization of an Octahedral Coordination Complex Inside a Self-Assembled Exoskeleton. Chem. - Eur. J. 2016, 22, 10791−10795. (33) Löffler, S.; Lübben, J.; Wuttke, A.; Mata, R. A.; John, M.; Dittrich, B.; Clever, G. H. Internal Dynamics and Guest Binding of a Sterically Overcrowded Host. Chem. Sci. 2016, 7, 4676−4684. (34) Krick, M.; Holstein, J.; Würtele, C.; Clever, G. H. Endohedral Dynamics of Push-Pull Rotor-Functionalized Cages. Chem. Commun. 2016, 52, 10411−10414. (35) Han, M.; Michel, R.; He, B.; Chen, Y.-S.; Stalke, D.; John, M.; Clever, G. H. Light-Triggered Guest Uptake and Release by a Photochromic Coordination Cage. Angew. Chem., Int. Ed. 2013, 52, 1319−1323. (36) Han, M.; Luo, Y.; Damaschke, B.; Gómez, L.; Ribas, X.; Jose, A.; Peretzki, P.; Seibt, M.; Clever, G. H. Light-Controlled Interconversion Between a Self-Assembled Triangle and a Rhombicuboctahedral Sphere. Angew. Chem., Int. Ed. 2016, 55, 445−449. (37) Harris, K.; Fujita, D.; Fujita, M. Giant Hollow MnL2n Spherical Complexes: Structure, Functionalisation and Applications. Chem. Commun. 2013, 49, 6703−6712. (38) Fukuda, M.; Sekiya, R.; Kuroda, R. A Quadruply Stranded Metallohelicate and Its Spontaneous Dimerization Into an Interlocked Metallohelicate. Angew. Chem., Int. Ed. 2008, 47, 706−710. (39) Freye, S.; Hey, J.; Torras Galán, A.; Stalke, D.; Herbst Irmer, R.; John, M.; Clever, G. H. Allosteric Binding of Halide Anions by a New Dimeric Interpenetrated Coordination Cage. Angew. Chem., Int. Ed. 2012, 51, 2191−2194. (40) Freye, S.; Engelhard, D. M.; John, M.; Clever, G. H. Counter Ion Dynamics in an Interpenetrated Coordination Cage Capable of Dissolving AgCl. Chem. - Eur. J. 2013, 19, 2114−2121. (41) Dieterich, J. M.; Clever, G. H.; Mata, R. A. A Push-and-Pull Model for Allosteric Anion Binding in Cage Complexes. Phys. Chem. Chem. Phys. 2012, 14, 12746−12749. (42) Freye, S.; Michel, R.; Stalke, D.; Pawliczek, M.; Frauendorf, H.; Clever, G. H. Template Control Over Dimerization and Guest Selectivity of Interpenetrated Coordination Cages. J. Am. Chem. Soc. 2013, 135, 8476−8479. (43) Löffler, S.; Lübben, J.; Krause, L.; Stalke, D.; Dittrich, B.; Clever, G. H. Triggered Exchange of Anionic for Neutral Guests Inside a Cationic Coordination Cage. J. Am. Chem. Soc. 2015, 137, 1060−1063. (44) Zhu, R.; Lübben, J.; Dittrich, B.; Clever, G. H. Stepwise HalideTriggered Double and Triple Catenation of Self-Assembled Coordination Cages. Angew. Chem., Int. Ed. 2015, 54, 2796−2800. (45) Frank, M.; Hey, J.; Balcioglu, I.; Chen, Y.-S.; Stalke, D.; Suenobu, T.; Fukuzumi, S.; Frauendorf, H.; Clever, G. H. Assembly and Stepwise Oxidation of Interpenetrated Coordination Cages Based on Phenothiazine. Angew. Chem., Int. Ed. 2013, 52, 10102−10106. (46) Croué, V.; Goeb, S.; Sallé, M. Metal-Driven Self-Assembly: the Case of Redox-Active Discrete Architectures. Chem. Commun. 2015, 51, 7275−7289. (47) Frank, M.; Ahrens, J.; Bejenke, I.; Krick, M.; Schwarzer, D.; Clever, G. H. Light-Induced Charge Separation in Densely Packed Donor-Acceptor Coordination Cages. J. Am. Chem. Soc. 2016, 138, 8279−8287. (48) Safont-Sempere, M. M.; Fernández, G.; Würthner, F. Self-Sorting Phenomena in Complex Supramolecular Systems. Chem. Rev. 2011, 111, 5784−5814. (49) Lal Saha, M.; Schmittel, M. Degree of Molecular Self-Sorting in Multicomponent Systems. Org. Biomol. Chem. 2012, 10, 4651−4684. (50) Mukherjee, S.; Mukherjee, P. S. Template-Free Multicomponent Coordination-Driven Self-Assembly of Pd(II)/Pt(II) Molecular Cages. Chem. Commun. 2014, 50, 2239−2248. (51) He, Z.; Jiang, W.; Schalley, C. A. Integrative Self-Sorting: a Versatile Strategy for the Construction of Complex Supramolecular Architecture. Chem. Soc. Rev. 2015, 44, 779−789. (52) Bloch, W. M.; Abe, Y.; Holstein, J. J.; Wandtke, C. M.; Dittrich, B.; Clever, G. H. Geometric Complementarity in Assembly and Guest Recognition of a Bent Heteroleptic Cis-[Pd2LA2LB2] Coordination Cage. J. Am. Chem. Soc. 2016, 138, 13750−13755. 2242


Article

Accounts of Chemical Research (53) Bloch, W. M.; Holstein, J. J.; Hiller, W.; Clever, G. H. Morphological Control of Heteroleptic cis- and trans-Pd2L2L′2 Cages. Angew. Chem., Int. Ed. 2017, 56, 8285−8289. (54) Bloch, W. M.; Clever, G. H. Integrative self-sorting of coordination cages based on ’naked’ metal ions. Chem. Commun. 2017, 53, 8506. (55) Klein, C.; Gütz, C.; Bogner, M.; Topić, F.; Rissanen, K.; Lützen, A. A New Structural Motif for an Enantiomerically Pure Metallosupramolecular Pd4L8 Aggregate by Anion Templating. Angew. Chem., Int. Ed. 2014, 53, 3739−3742.

2243


CationâAnion Arrangement Patterns in Self-Assembled Pd2L4 and

Recommend Documents

CationâAnion Arrangement Patterns in Self-Assembled Pd2L4 and