Number of Channels Discrete or Continuous

1. Introduction

Section:

In biological systems, nano- and angstrom-spaces and voids generated between molecules express various functions depending on their sizes and dimensions. In particular, one-dimensional (1D) channel structures, which are constructed by molecular assembly, play important roles in biological systems.^{1 – 6} Such 1D channel structures have spaces of certain lengths, therefore molecules can go into the channel at the entrance and leave via the exit. It takes time to pass through the channel from the entrance to the exit. These features of 1D channels contribute to the sophisticated functions observed in biological systems. For example, microtubes, which consist of tubulin dimers and have several nanometer-scale channels in their structures, act as transporters in nerve cells.^{1 – 4} Ion channels are located in cell membranes, and control the exchange of ions between outer and inner cells.^{5 , 6} The mimicking of these biological systems, by controlling the lengths and diameters of 1D channels using synthetic molecules is an important research target.⁷

Pillar[n]arenes (Figure 1 ), which were first reported by our group in 2008, are comparatively new macrocyclic compounds in supramolecular chemistry.^{8 – 10} One of their important aspects is their shape. Because 1,4-dialkoxybenzene units are connected by methylene bridges at the 2,5-position (para-position), their structures are highly symmetrical polygons, i.e., pillar[5]arenes and pillar[6]arenes are pentagons and hexagons, respectively. Because of their highly symmetrical structures, they are suitable building blocks for forming 1D channel structures. For example, continuous non-covalent 1D channels can be obtained by connecting both pillar[n]arene rims via physical interactions. Another process for constructing non-covalent 1D channels is formation of poly(pseudo)rotaxane structures. Included linear polymeric chains can be triggered to form 1D channels. Discrete 1D channel structures have recently been developed by using rim-differentiated pillar[n]arenes. Rim-differentiated pillar[5]arenes have differing rims, therefore, by introducing an interaction site on one rim and a non-interaction site on another rim, discrete channel assemblies, i.e., tubular dimers and trimers can be obtained. Continuous and discrete 1D channels can also be produced by connecting pillar[n]arene units via covalent bonds. In this mini-review, we will discuss the use of non-covalent interactions and covalent bonds for the preparation of continuous and discrete 1D channels. We will also discuss the characteristic features and potential uses of these 1D channels in material applications.

2. Non-Covalent 1D Channels Based on Pillar[n]arenes

Section:

The symmetrically polygonal structures and ease of functionalization of pillar[n]arenes indicate their potential use in constructing 1D channels through multiple non-covalent interactions by decorating their rims. Intermolecular non-covalent interactions such as hydrogen bonding and ionic interactions in the axial direction enable the formation of channel assemblies of pillar[n]arenes. Generally, mixing pillar[n]arenes with identical decorations on both rims leads to formation of 1D channels of uncontrollable length. This is because continuous intermolecular interactions occur in such systems. In this section, we will focus on the 1D channels produced by non-covalent interactions in the axial direction of pillar[n]arenes and their further assembly. Recent efforts to control the lengths of 1D channel structures formed from pillar[n]arenes will also be addressed.

2.1 Continuous 1D Channels Constructed via Non-Covalent Interaction of Pillar[n]arenes.

We reported the first example of channel assembly of pillar[n]arenes, based on per-hydroxy-pillar[5]arene H1 (Figure 2 ).¹¹ H1 was initially dissolved in acetone, a polar solvent, which can interact with the hydroxy groups on the rims of H1 via hydrogen bonding. On increasing the proportion of chloroform, a relatively nonpolar solvent, a violet precipitate appeared. The structure of this precipitate was investigated by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Uniform channel structures of average width approximately 150 nm were observed. These observations suggest that intermolecular hydrogen bonding of hydroxy groups triggered the formation of a channel assembly. In contrast, guest inclusion with octylviologen V-C8 prevented precipitation because of disruption of hydrogen bonds between neighboring H1 units.

Hou and coworkers reported gelation of H1 in a mixture of organic solvents.¹² Stable (up to one month) organogels with the minimum gelation concentration at 1.3 mg/mL were obtained. Fibers were observed by both SEM and TEM. Field-emission TEM images showed that the fibers were composed of 1D channels. X-ray diffraction (XRD) showed repeated stacking of H1 units. Clearly, the channel assembly was initially formed. Then, the constructed 1D channels aggregated to generate fibers, which were stabilized by aromatic stacking, and supramolecular organogels were produced.

Similarly, Cohen and coworkers reported rapid gelation of a mixture of decacarboxylate-pillar[5]arene H2 with decaamino-pillar[5]arenes H3a–c at low concentrations (<1% w/v) in various solvents (Figure 3 ).¹³ The two-component supramolecular organogels were shape-based complementary, which means that neither monomeric diaminobenzene H4 nor dodecaamino-pillar[6]arene H5 gelated in the presence of H2. In both cases, precipitation was observed. These results strongly suggest that 1D channel assembly between H2 and H3 occurred through multiple charge-assisted hydrogen bonds in the constructed organogels. Furthermore, they found that the gelation of these two-component organogels based on acidic and amino pillar[n]arenes proceeded via shape-directed sorting in mixtures of pillar[5]arenes and pillar[6]arenes. Only acidic pillar[n]arenes and amino-pillar[n]arenes of the same size could generate 1D channels and gels.¹⁴

2.2 Length-Controllable Discrete 1D Channels.

Unlike the case for continuous 1D channels, length control of self-assembled 1D channels based on pillar[n]arenes is difficult. Termination of continuous interactions requires the use of pillar[n]arenes with different functionalizations of the two rims. New procedures for synthesizing rim-differentiated pillar[5]arenes, which were reported by Ma et al.¹⁵ and Zuilhof and Sue et al.,¹⁶ have enabled the construction of length-controlled 1D channels. We designed and synthesized the rim-differentiated acidic pillar[5]arene H6. The benzoic acid group on one rim of H6 guarantees assembly via hydrogen bonds in the axial direction to form a channel assembly. The alkyl group on the other rim increases the solubility of H6 in organic solvents and stops continuous assembly of the molecule. We successfully synthesized discrete 1D channels with clear structures and controllable lengths via intermolecular hydrogen bonding on the molecular axis (Figure 4 ).¹⁷ In chloroform, H6 dimerized with increasing concentration or on addition of dibromo linear guests. In the same solvent, mixing H6 and peramino-pillar[5]arene H7 in a 1:2 ratio gave trimeric 1D channel assemblies. A guest molecule, namely 1,4-dibromobutane (DBB), in these confined cavities (i.e., monomer, dimer, and trimer) could be controllably exchanged with the outside. The release rate was closely related to the length of the 1D channels (Figure 4 ). In addition, the dimer (H6)₂ shows structural stability and distinct pore shrinkage [from a diameter of ca. 5 Å in monomeric H6 to 2.8 Å in (H6)₂]. Collaboration between our and Barboiu's groups enabled the successful use of (H6)₂ to simulate water channels and achieve exclusive and highly selective transport of water molecules across a membrane.¹⁸ This has significance for achieving an in-depth understanding of the aquaporin mechanism. We also produced two types of chiral 1D channels, i.e., left-handed and right-handed ones, by introducing stereogenic carbons into pillar[5]arene substituents.¹⁹ These chiral 1D channels could be used for chiral molecule transportation and separation.

3. Polyrotaxanes and Polypseudorotaxanes: 1D Channels Penetrated by Linear Polymers

Section:

Pillar[n]arenes form host–guest complexes not only with small molecules but also with linear polymers. When many pillar[n]arene macrocycles are threaded onto a polymer axle, the macrocycles are aligned along the polymer to afford a 1D channels under topological confinement. This section describes such templated synthesis of channel assemblies of pillar[n]arenes and the unique properties of the obtained poly(pseudo)rotaxanes. The characteristics, which arise from the dynamic assembled structures and interactions with axle components, make threading into pillar[n]arenes an easy and versatile strategy for altering the nature of polymers.

3.1 Poly(pseudo)rotaxanes Based on Electron-Accepting Polymeric Chains.

In the preparation of (pseudo)rotaxanes, stable complexation between the axle and ring components is critically important in enabling the washing process and efficient capping. Pillar[n]arenes consist of electron-rich hydroquinone units and therefore strongly capture electron-accepting species such as pyridinium cation derivatives.⁸ This feature was first used by us in 2010 for the efficient preparation of polypseudorotaxanes²⁰ and a polyrotaxane²¹ constructed from pillar[5]arene and viologen polymers (Figure 5 ). When phenolic pillar[5]arene H1 was mixed with a viologen polymer with a C8 linker (PV-C8) in acetone-d ₆/acetonitrile-d ₃ (1/1, v/v), the threaded structure polypseudorotaxane 1 was formed. In contrast, a threaded structure was hardly formed when a viologen polymer with a C3 linker (PV-C3) was used. This is because short alkyl chains do not provide enough space for stable complexation. The shuttling of H1 in polypseudorotaxane 1 was solvent dependent: it was fast in acetone-d ₆/acetonitrile-d ₃ and slow in DMSO-d ₆. Heating at 60 °C speeded up the movement in DMSO-d ₆ and dethreading was observed above 110 °C. It is worth noting that host–guest complexation was completely prevented by capping both ends of PV-C8 with adamantane-derived bulky groups in advance.

On the basis of this finding, we used adamantane units as stoppers for a polyrotaxane constructed from H1 and PV-C8. Excess 1-adamantyl bromomethyl ketone was added to a mixture of H1 and PV-C8 in DMF/acetonitrile (1/4, v/v). Copious washing with acetonitrile provided the target polyrotaxane 2 in 93% yield. The high yield was achieved because of the high stability of the threaded structure in DMF/acetonitrile, even at 100 °C. The introduction of stopper units greatly altered the properties of the supramolecular complexes. The components H1 and PV-C8 are soluble in acetone, acetonitrile, methanol, DMF, and DMSO, but polyrotaxane 2 can be dissolved only in DMF and DMSO. The reduced solubility and different assembled structure are ascribed to intermolecular hydrogen bonds. This was confirmed by a clear shift in the Fourier-transform infrared (FT-IR) spectrum. When a solution of polyrotaxane 2 in DMSO-d ₆ was heated, the polymeric chain signals were almost unchanged and OH peaks of H1 were up-field-shifted in the ¹H NMR spectrum. These observations suggest that shuttling was activated without dethreading under topological confinement. Heating intensified the clear absorption peak at 540 nm in the ultraviolet-visible absorption spectrum of 2 in DMSO. Upon heating, the shuttling of H1 became fast, which contributes efficient viologen radical cation formation by electron transfer from H1 to the viologen units. The efficient radical cation formation results in the solution color changing from yellow to violet. The spectrum of polypseudorotaxane 1 also had a broad shoulder band at 400–600 nm, which is assigned to a charge-transfer complex, but the band intensity decreased on heating because of dethreading. This is in sharp contrast to the case for polyrotaxane 2.

In 2011, our group reported the threading of H1 onto a polyaniline to form a polypseudorotaxane and concomitant reduction of the oxidized units in polyaniline (Figure 6 ).²² Polyaniline is a well-known conductive polymer that shows vivid colors, depending on its redox state. These states are reversibly interconvertible with each other between the dark-violet fully oxidized form (pernigraniline base, PB), blue half-oxidized one (emeraldine base, EB), and transparent fully reduced one (leucoemeraldine base, LB). First, H1 was added to a blue solution of the EB form of polyaniline. The blue solution became transparent, which was interpreted as reduction of the EB units to LB ones.²³ This color change was accompanied by the emergence of an emission band at 416 nm because of the absence of quinoidal segments, and a new FT-IR spectroscopic peak at 1646 cm⁻¹, which is assigned to C=O stretching of the benzoquinone units in pillar[5]arenes. Addition of the per-methylated derivative H8 to the EB form of polyaniline instead of H1 did not induce a similar color change. These results indicate that the hydroquinone units in H1 reduced the EB segments of polyaniline to yield benzoquinone units and the LB form. Addition of simple hydroquinone did not change the color much. The enhanced reducing ability of H1 shows that the reaction can be promoted inside a 1D channel structure.

3.2 Poly(pseudo)rotaxanes Based on Neutral Polymeric Chains.

Linear polymers consisting of highly electron-deficient units are effective building blocks for poly(pseudo)rotaxanes because of their strong complexation with cyclic hosts. However, such topological assemblies tend to be poorly soluble in non-polar organic solvents and their extension to high-molecular-weight systems is difficult.^{20 , 21} These factors decrease their utility as materials. Efficient protocols for producing polyrotaxanes from non-cationic polymers are therefore needed.

To meet this demand, we developed poly(tetrahydrofuran)-based polyrotaxanes by using pillar[5]arenes (Figure 7 );²⁴ their host–guest chemistry is not limited to strongly electron-deficient species but is applicable to alkyl chains because of multiple CH–π interactions. To maximize the reaction concentration, solid pillar[5]arenes were altered to liquid ones in advance by functionalization with tri(ethylene oxide) segments.²⁵ Under typical solution conditions (¹H NMR in CDCl₃), azido-terminated poly(tetrahydrofuran) (PTHF) did not form a stable polypseudorotaxane with liquid pillar[5]arene H9. In contrast, in a bulk system (without CDCl₃), the predominant product was a polypseudorotaxane. Polyrotaxane 4 was then synthesized under neat conditions by capping the azido terminals with alkyne-appended stoppers under copper catalysis. Under typical conditions with CHCl₃ as the solvent, polyrotaxane 4 (average cover number: 7) was obtained but the conversion (22%) was low because of weak complexation. In contrast, the use of neat conditions with excess liquid H9 (10 equiv per polymer unit) greatly increased the conversion and average cover number to 71% and 27, respectively; this enabled product isolation in 44% yield. These results indicate that polyrotaxane synthesis proceeds better under neat conditions than under typical conditions with solvents. Polyrotaxane synthesis was also achieved by using liquid pillar[5]arene H10, which bears vinyl groups, instead of H9. A vinyl-bearing polyrotaxane 5 (average cover number: 12) was obtained in 35% isolated yield and then converted to a topological gel by alkene metathesis with a Grubbs catalyst. The gel was a chemical gel that swelled in organic solvents with intermediate dielectric constants, e.g., CHCl₃, CH₂Cl₂, and THF, but did not respond to non-polar hexane or highly polar acetone and methanol.

The use of melt polymers and cyclic hosts is another effective method for avoiding preparation in dilute solutions. In 2014, we reported that melt mixing of polyethylene (PE) and alkoxy-substituted pillar[5]arene H11 ²⁶ yielded polypseudorotaxanes with altered assembled states (Figure 8 ).²⁷ First, complexation of H11 with eicosane (model compound for PE) was performed in typical solvent and bulk systems. In a solvent system (CDCl₃), the association constant of the complex was estimated by ¹H NMR spectroscopy to be K = 29 M⁻¹, with 1:1 stoichiometry. This indicates that the complex is not very stable in solution. Then, complexation in the bulk system was investigated. When H11 (mp 83 °C) was added to liquid eicosane (mp 36 °C) at 110 °C without a solvent, the mixture changed into a solid. Differential scanning calorimetry (DSC) studies showed that the intensities of the endothermic peaks for eicosane melting at 36 °C and H11 at 83 °C decreased significantly and a new peak emerged at 135 °C. These results indicate that bulk complexation is more efficient than that in solution. Based on the model experiment using eicosane, H11 was mixed with high-density PE (HD-PE, mp 126 °C) in the molten state at 140 °C; this caused similar solidification. The DSC curves contained a new endothermic peak at 152 °C and the intensities of the peaks for pristine HD-PE and H11 decreased. The optimum mixing ratio was determined to be HD-PE/H11 = 30/70 (w/w). Similar molten-state complexation was observed with low-density PE (LD-PE) and ultra-HD-PE, but not with polypropylene. A polypseudorotaxane with a LD-PE axle gave a broad DSC peak at 130–140 °C, with a peak for pristine LD-PE at 92 °C, and that with ultra-HD-PE gave a peak at 152 °C. The assembled structures of a HD-PE-based polypseudorotaxane and its components were investigated by using powder XRD, small-angle X-ray scattering, and tapping-mode atomic force microscopy. On addition of H11, the lamellar structure of HD-PE changed to an amorphous phase because of polypseudorotaxane formation. The observed lamellar-to-amorphous phase change on polypseudorotaxane formation was reversed by addition of a competitive guest, i.e., DBB. These chemically responsive transitions between the solid and molten states provide a unique method for modulating the properties of HD-PE.

In 2015, these polypseudorotaxanes constructed from PE and pillar[5]arenes were extended to block copolymers of PE and poly(ethylene oxide) (PEO) by Li and coworkers.²⁸ When the block copolymer b-PE-PEO ₁₄₀₀ was mixed with alkoxy-substituted pillar[5]arene H12, H12 formed polypseudorotaxane 7 and was localized on PE chains with high site selectivity. In addition, the introduction of competitive guests, i.e., DBB or 1,4-dicyanobutane (DCB), could disassemble the complex, as shown by ¹H NMR titration studies. The chemical response caused a decrease in the solubility in CDCl₃ and a turbid mixture formed; this is consistent with the lower solubility of free b-PE-PEO ₁₄₀₀ compared with that of polypseudorotaxane 7. The aggregates were examined by SEM and TEM. The free copolymers b-PE-PEO ₁₄₀₀ and b-PE-PEO ₂₂₅₀ aggregated into irregular and island-shaped particles, respectively, and polypseudorotaxane 7 formed network structures. The higher regularity in b-PE-PEO ₂₂₅₀-based materials can be attributed to longer unthreaded PEO chains, which were softer and more adaptive than the threaded PE ones. The addition of competitive guests, i.e., DBB or DCB, destroyed the network and square-plate or irregular aggregates, respectively, were formed because of dethreading.

The above results suggest that PEO is a weak guest for pillar[n]arenes, therefore it is difficult to use it to prepare polypseudorotaxanes and polyrotaxanes. When a PEO with OH ends (PEO_1kOH) was mixed with H12 in CDCl₃, neither ¹H NMR peak shifts nor new peaks were detected. Under these conditions, simple linear n-alkanes cause spectral changes and negatively charged oxygen atoms are thought to disturb PEO threading into the cavity. To overcome this problem, we developed a heterogeneous method. The pre-activated crystalline pillar[5]arene H12, which did not contain guest solvents, was immersed in melted PEO_1kOH at 80 °C (Figure 9 ).²⁹

After filtration and washing with water to remove uncomplexed PEO_1kOH, the residue was dissolved in CDCl₃ and investigated by ¹H NMR spectroscopy. In addition to signals from H12, peaks from the methylene protons of PEO were clearly observed; this suggests PEO_1kOH uptake at an ethylene oxide (EO) unit/H12 molar ratio of ca. 4.0. Complexation in the solid state was confirmed by solid-state magic-angle-spinning 2D hetero-correlated NMR spectroscopy. The disappearance of the endothermic peaks at 40 and 156 °C for pristine PEO and H12, respectively, and the appearance of a new peak at 179 °C in the DSC curve also indicated a threaded structure. In these 1D channels, the PEO dynamics are greatly restricted compared with those of uncomplexed PEO. This was confirmed by the half-width at half-maximum values for the solid-state ¹³C NMR spectroscopic peaks of PEO.

Crystals of H12 preferentially take up high-molecular-weight PEO. Longer PEO_10kOH replaced shorter PEO_1kOH captured in H12 crystals on re-immersion, but the reverse replacement did not occur. Liquid chromatography showed that when activated H12 crystals were immersed in an equal-weight mixture of PEO_1kOH, PEO_4kOH, PEO_6kOH, and PEO_10kOH, the PEOs with high molecular mass fractions were selectively taken up. The PEO terminals also affected the rate of complexation with H12 crystals. Non-hydrogen-bonding PEO_1kOMe reached equilibrium in 3 min, but PEO_1kOH and PEO_1kNH₂ took longer (10 and 20 min, respectively). Complexation of bulky PEO_1kOTs was also slow. In the case of PEO_1kCO₂H, the PEO uptake decreased to the EO/H12 value = ca. 1.0, and the saturation time increased significantly.

We recently found that hydrophobic effects and hydrogen bonds can be used in the preparation of PEO-based polypseudorotaxanes (Figure 10 ), instead of highly concentrated conditions that involve liquid hosts, melt mixing, and crystal immersion.³⁰

Phenolic pillar[5]arene H1, which is highly soluble in methanol, was used as a ring component, and gave a precipitate of polypseudorotaxane 8 in 32% yield via complexation in methanol/water (1/1, v/v). The EO/H1 value was determined to be 8 by ¹H NMR spectroscopy in acetone-d ₆ and the threaded structure was confirmed by solid-state ¹H/¹³C NMR spectroscopy. Control experiments under basic conditions with NaOH and in the absence of water supported the importance of the hydrophobic effect and hydrogen bonds. Changing the amount of added PEO altered the EO/H1 value from 8 to 23, and this method was successfully applied to PEOs with large molecular weights, up to 500 000, and PEOs bearing various terminal groups. The COOH terminals of polypseudorotaxane 8 were capped with 1-adamantylamine in acetone to produce polyrotaxane 9 (EO/H1 = 9) in 43% yield. This reaction did not proceed with polypseudorotaxane 8 bearing per-ethylated H12 rings and the number of ring units decreased after reaction in polar solvents such as acetonitrile (EO/H1 = 34). These results indicate the significance of H1 hydrogen bonds in preventing dethreading, which competes with the capping reaction. A high-molecular-weight axle was also used in polyrotaxane synthesis to produce polyrotaxane 10 (EO/H1 = 28). The materials did not melt above 60 °C but became elastic with shape-memory behavior, while retaining a solid state, because of cross-linked hydrogen bonding with H1.

4. 1D Channel Architectures Based on Covalently Linked Pillar[n]arenes

Section:

1D channel architectures based on pillar[n]arenes have also been built via covalent linkages. This approach is beneficial for constructing well-controlled and robust 1D channel structures. In particular, the binding abilities and selectivities toward guest molecules can be altered because of their cooperative binding systems based on multiple pillar[n]arenes cavities. In this section, we will discuss the functionalities of covalently linked pillar[n]arene dimeric, oligomeric, and polymeric systems.

4.1 Dimeric and Oligomeric Systems.

We synthesized pillar[5]arene dimer H13 by connecting two mono-hydroxy pillar[5]arenes with 1,4-bis(bromomethyl)benzene (Figure 11 a).³¹ Because of the connected π-electron-rich spaces in H13, its ability to bind guest molecules is better than that of the monomeric pillar[5]arene H8. The π-electron-rich cavities in the pillar[5]arene enhanced CH/π hydrogen bonding, and the association constant of the 1:1 H13–n-hexane complex was 98 ± 12 M⁻¹. The binding ability of monomeric pillar[5]arene H8 toward n-alkanes was too weak to determine the stoichiometry and association constants. The dimer H13 showed high selectivity only for linear n-alkanes. Bulky cyclohexane and branched 2- and 3-methylhexane hardly interacted with H13, whereas the compact linear shape of n-alkanes is favorable for complexation. This is because of the narrow cavity size and well-defined channel architecture of pillar[5]arenes.

The effects of bridging linkers on guest binding was investigated by Jia and coworkers.³² They synthesized three pillar[5]arene dimers, which were bridged with a relatively rigid phenylene unit (H13 and H14) or a flexible aliphatic chain (H15), and evaluated the linker effects (Figure 11 a). The binding affinity of H15 toward the aliphatic dibromide guest C10-2Br is stronger than those of the rigid dimers H13 and H14. Dimer H15 can adopt suitable conformations for deep guest penetration across two pillar[5]arene cavities because of its flexible methylene linker. This leads to strong hydrogen bonding and CH/π interactions between the host and guest. It should be noted that the association constant of the C10-2Br@H15 complex is higher by a factor of 18 than that of the complex between C10-2Br and monomer H8. These results suggest that the two pillar[5]arene cavities can bind the guest molecule cooperatively. The binding between H13, which has a relatively rigid long 1,4-phenylene linker, and a bromo-substituted pyromellitic diimide (PDI-2Br) guest, was stronger than those of the other two dimers. In addition to CH/Br interactions, the enhanced binding ability originates from π–π interactions between the 1,4-phenylene linker and PDI part of the guest. This further promotes host–guest inclusion complexation (Figure 11 b).

Hou and coworkers reported that self-assembled pillar[5]arene channel structures that contain 10 ester groups can provide a pathway for proton transport in crystals and operate water-wire-based artificial proton channels because water molecules form ordered water wires inside the cavities.³³ They investigated the possibility of forming transmembrane proton channels by incorporating the decaester pillar[5]arene monomer H16 into lipid bilayers (Figure 12 a). However, the thickness (≈3.7 nm) of the hydrophobic part of the lipid bilayer is more than double the length of the monomer H16 (1.6 nm). Formation of a channel that matches the thickness of the bilayer system therefore needs face-to-face stacking of two pillar[5]arene monomers during self-assembly. Hou's group therefore designed dimeric decaester pillar[5]arenes with linkers of different lengths. The hexamethylene linker in H21 enhanced the intramolecular stacking interactions of the pillar[5]arene moieties and gave a three-fold increase in the probability of the channel being open compared with that in the case of monomer H16 (Figure 12 b).

This result indicates that the channel was more stable than the assembled dimeric channel of monomer H16, therefore the resulting stable water wire in the channel of H19 gave the highest proton conductance. The other four dimers showed no activity or were less efficient than H19, which implies that their linkers could not achieve good intramolecular dimeric stacking for proton channels (Figure 12 c).

Stoddart and coworkers synthesized 1D tubular dimers and higher oligomers by forming oxazoles on an amino-functionalized pillar[5]arene, in which the planes of the pillar[5]arene rings are orthogonal to the conjugated backbone of the array (Figure 13 ).³⁴ The dimer H23 formed a 1:1 complex with an extended bypridinium dication, Ex-Bpy. The association constant for this complex is (2.16 ± 0.47) × 10³ M⁻¹, which is comparable to those of analogous complexes between pillar[5]arenes and bipyridinium derivatives, despite the increase in the electron density of the bipyridinium units of EX-Bpy as a result of the extended nature of the guest.

H22 can be used as an AB-type monomer for constructing rigid oligomers with benzodioxazole linkages formed orthogonally to the pillar[5]arene rings. Polycondensation reactions gave the desired oligomers H24–H31, which ranged from two (2-Mer) up to nine (9-Mer) repeating units. Theoretical calculations on the 4-Mer and 9-Mer species indicate that these extended structures can be as long as 5 and 11 nm, respectively. These unique dimeric and oligomeric systems based on pillar[5]arenes represent novel building blocks in supramolecular chemistry.

4.2 Polymeric System.

Müllen and coworkers prepared a conjugated polymer in which one of the benzenes in the pillar[5]arene rings is incorporated into a poly(arylene ethynylene) backbone.³⁵ A Sonogashira–Hagihara coupling reaction gave the target polymer 11 with a relative number-average molecular weight of approximately 16 kDa and a polydispersity index of 3.82 (Figure 14 a). The conjugated polymer can be regarded as a fishing-rod-like conjugated hydrocarbon 1D channel in which the conjugated backbone is the rod and the pillar[5]arene units are the guides arranged along the underside of the fishing rod, in which a 1D channel structure can be formed.

We synthesized a similar rod-like conjugated polymer with a phenol-containing pillar[5]arene in its main chain (Figure 14 f).³⁶ The phenolic polymer 13 was prepared by deprotection with BBr₃ of the ethoxy groups of the ethyl moieties in the ethoxy-pillar[5]arene conjugated polymer 12 (Figure 14 b). Because of its thermally stable poly(p-phenylene) main chain, the thermal stability of the phenolic polymer 13, which has a 10 wt % loss temperature (T ₁₀) of 300 °C, is higher than that of the corresponding per-ethylated pillar[5]arene H12 (164 °C).

The phenolic polymer 13 showed selective gas adsorption properties. It selectively adsorbed CO₂, but did not uptake N₂ or methane. Intermolecular hydrogen bonds can be formed between the phenolic moieties in 13. This gives it a rigid and controlled porous structure. The included CO₂ was hardly released during desorption, which indicates that it was stably captured by the porous structure of 13. Similarly, a pillar[5]arene with 10 phenolic moieties can selectively adsorb CO₂ because of its molecular-scale porosity and the dipole interactions between CO₂ and the phenolic moieties on the pillar[5]arene core. This indicates that the intrinsic CO₂ adsorption properties of the phenolic pillar[5]arene monomer are retained in the polymeric compound.

Cao and coworkers synthesized conjugated pillar[5]arene–diketopyrrolopyrrole (DPP) copolymer 14 (Figure 14 c). They reported that its selectivity for neutral guests is higher than that of the monomeric pillar[5]arene.³⁷ Polymer 14 was able to form a linear main-chain polypseudorotaxane and showed strong host–guest binding affinities toward DCB but low binding affinities toward 1,4-dihalobutanes (DClB, DBB, and DIB) and 1,4-bis(imidazol-1-yl)butane (BIB), to which well-known pillar[5]arene derivatives bind strongly (Figure 14 e). This suggests that the selectivity of pillar[5]arene guest binding can be improved by polymerization. The specific binding ability of 14 can be used for adsorbing organic pollutants in water. A comparison of the integral proportions in the ¹H NMR spectra of DCB in D₂O before and after addition of solid 14 showed that 14 adsorbed DCB from the aqueous phase almost completely in 30 h. This indicates that 14 is a promising material for adsorbing and separating toxic pollutants from water.

Wen and coworkers showed that a polymer containing pillar[5]arenes in the poly(p-phenylene terephthalamide) (PPTA) polymer 15 chain can be used to adsorb organic pollutants in water (Figure 14 d).³⁸ A ¹H NMR spectroscopic study showed that polymer 15 has an exceptional ability for adsorbing short-chain alkyl guests (amines, alcohols, and carboxylic acids) in water. Removal efficiencies (RE: the percentage reduction in concentration of the compounds) from 47.5% to 98.6% were achieved for the tested short-chain alkyl guests in D₂O solution. Dimethoxy-pillar[5]arene H8 adsorbed n-hexanoic acid, n-hexylamine, and n-hexanol, but the RE values were only 17%, 25%, and 6%, respectively. This superiority of polymer 15 over unmodified pillar[5]arenes can be attributed to its hybrid structure of pillar[5]arene rings and PPTA chains, in which the PPTA chains serve as skeletons to support pillar[5]arene struts, which have cavities that are wide open to short-chain intruders. PPTA is a typical wholly aromatic polyamide and has a wide range of industrial applications because of its excellent mechanical properties, and chemical and thermal stabilities. Wen's group therefore investigated the adsorbent's reusability, which is a key factor in water and wastewater treatment. After reactivation, polymer 15 was used in the treatment of D₂O solutions of n-hexanoic acid, n-hexylamine, or n-hexanol for five successive cycles without significant loss of RE. This unambiguously shows the robustness of this absorbent material. These results show that pillar[5]arene-based polymers are a promising class of absorbent materials for removing linear short-chain aliphatic contaminants in water.

5. Conclusion

Section:

In this mini-review, we described the preparation of 1D channels by forming assemblies of pillar[n]arenes via non-covalent interactions or by connecting pillar[n]arenes via covalent bonds. First, continuous 1D channels were produced by introducing physical interaction sites on both rims. However, we could not achieve precise 1D channel length control by using pillar[n]arenes with interaction sites on both rims because connections between pillar[n]arenes occur continuously, and 1D channels were obtained as precipitates. The development of rim-differentiated pillar[n]arene procedures, enable the construction of length-controllable discrete 1D channels, dimers, and trimers. This is because rim-differentiated pillar[n]arenes, which have interaction sites on one rim and non-interaction sites on the other rim, stop continuous 1D channel formation. Precise control of the 1D channel length, showed that the transportation speed of guest molecules depends on the 1D channel length. A future target is therefore to control access of guest molecules into 1D channels by modification of stimuli-responsive gates into the entrance and exit rims.

We also synthesized continuous 1D channel structures by formation of poly(pseudo)rotaxanes. Polymeric chains act as templates for forming 1D channel structures. Length-controlled 1D channels could therefore be produced by controlling the length of the polymeric chain template.

By connecting single pillar[n]arene units via covalent bonds, we synthesized continuous 1D channels, and discrete dimers and trimers. However, these covalent 1D channel structures were produced by connecting single pillar[n]arene units. All unit-covalently linked 1D channel structures should be rigid 1D channel structures and have high stability, and this is therefore our next research target. Further development of efficient functionalization of pillar[n]arenes would enable construction of pillar[n]arene-based 1D channels with highly functional and sophisticated transportation systems.

This work was supported by the Grant-in-Aid for Scientific Research on Innovative Areas: π-System Figuration (JP15H00990 and JP17H05148, T.O.), Soft Crystal (JP18H04510 and JP20H04670, T.O.), Kiban A (JP19H00909, T.O.), Research Activity Start-up (JP20K22528, K.K.), and Early-Career Scientists (JP21K14611, K.K. and JP21K14612, S.F.) from MEXT, Japan, JST CREST (JPMJCR18R3, T.O.), and the World Premier International Research Center Initiative (WPI), MEXT, Japan.

Kenichi Kato

Kenichi Kato received his Ph.D. degree (2020) from Kyoto University (Supervisor: Prof. Atsuhiro Osuka). During his Ph.D. studies, he was a JSPS research fellow (DC1, 2017–2020). In 2020, he started his academic career as an assistant professor at Kyoto University. His research interests cover π-conjugated and supramolecular systems.

Shunsuke Ohtani

Shunsuke Ohtani received his Ph.D. degree (2021) from Kyoto University (Supervisor: Prof. Kazuo Tanaka). In 2018, he worked as a visiting research fellow in John R. Reynolds's group at Georgia Institute of Technology in the USA. He has been an assistant professor at Kyoto University since 2021. His research interests focus on polymeric and supramolecular materials.

Shixin Fa

Shixin Fa received his Ph.D. degree (2015) from the Institute of Chemistry, Chinese Academy of Sciences. After postdoctoral research (2015–2018) at Iowa State University, he joined the WPI Nano Life Science Institute at Kanazawa University in 2018. He moved to Kyoto University in June 2019, where he worked with Prof. Tomoki Ogoshi on a CREST project. His current research interest is supramolecular assembly of pillar[n]arenes.

Tomoki Ogoshi

Tomoki Ogoshi received his Ph.D. degree (2005) from Kyoto University (Supervisor: Prof. Yoshiki Chujo). He was a JSPS postdoctoral research fellow (2005–2006) at Osaka University (Prof. Akira Harada). He joined Kanazawa University, where he was promoted to assistant professor in 2006, associate professor in 2010, and full professor in 2015. In 2019, he moved to Kyoto University. His research interests include supramolecular materials.

1)	L. A. Amos, T. S. Baker, Nature 1979, 279, 607. Crossref, Medline, CAS
2)	R. F. Luduena, Mol. Biol. Cell 1993, 4, 445. Crossref, Medline, CAS
3)	K. F. Sullivan, Annu. Rev. Cell Biol. 1988, 4, 687. Crossref, Medline, CAS
4)	R. H. Wade, D. J. Chretien, J. Struct. Biol. 1993, 110, 1. Crossref, Medline, CAS
5)	D. A. Doyle, J. Morais Cabral, R. A. Pfuetzner, A. Kuo, J. M. Gulbis, S. L. Cohen, B. T. Chait, R. MacKinnon, Science 1998, 280, 69. Crossref, Medline, CAS
6)	E. Tajkhorshid, P. Nollert, M. O. M. Jensen, L. J. W. Miercke, J. O'Connell, R. M. Stroud, K. Schulten, Science 2002, 296, 525. Crossref, Medline, CAS
7)	M. Barboiu, A. Gilles, Acc. Chem. Res. 2013, 46, 2814. Crossref, Medline, CAS
8)	T. Ogoshi, S. Kanai, S. Fujinami, T. Yamagishi, Y. Nakamoto, J. Am. Chem. Soc. 2008, 130, 5022. Crossref, Medline, CAS
9)	T. Ogoshi, T. Yamagishi, Y. Nakamoto, Chem. Rev. 2016, 116, 7937. Crossref, Medline, CAS
10)	M. Xue, Y. Yang, X. Chi, Z. Zhang, F. Huang, Acc. Chem. Res. 2012, 45, 1294. Crossref, Medline, CAS
11)	T. Aoki, T. Ogoshi, T. Yamagishi, Chem. Lett. 2011, 40, 795. Link, CAS
12)	W. Si, X.-B. Hu, X.-H. Liu, R. Fan, Z. Chen, L. Weng, J.-L. Hou, Tetrahedron Lett. 2011, 52, 2484. Crossref, CAS
13)	Y. Zafrani, D. Kaizerman, M. Hadar, N. Bigan, E. Granot, M. Ghosh, L. Adler-Abramovich, F. Patolsky, Y. Cohen, Chem.—Eur. J. 2018, 24, 15750. Crossref, Medline, CAS
14)	D. Kaizerman-Kane, M. Hadar, E. Granot, F. Patolsky, Y. Zafrani, Y. Cohen, Org. Chem. Front. 2019, 6, 3348. Crossref, CAS
15)	J. Ding, J. Chen, W. Mao, J. Huang, D. Ma, Org. Biomol. Chem. 2017, 15, 7894. Crossref, Medline, CAS
16)	M. Guo, X. Wang, C. Zhan, P. Demay-Drouhard, W. Li, K. Du, M. A. Olson, H. Zuilhof, A. C.-H. Sue, J. Am. Chem. Soc. 2018, 140, 74. Crossref, Medline, CAS
17)	S. Fa, Y. Sakata, S. Akine, T. Ogoshi, Angew. Chem., Int. Ed. 2020, 59, 9309. Crossref, Medline, CAS
18)	D. Strilets, S. Fa, A. Hardiagon, M. Baaden, T. Ogoshi, M. Barboiu, Angew. Chem., Int. Ed. 2020, 59, 23213. Crossref, Medline, CAS
19)	S. Fa, K. Adachi, Y. Nagata, K. Egami, K. Kato, T. Ogoshi, Chem. Sci. 2021, 12, 3483. Crossref, Medline, CAS
20)	T. Ogoshi, Y. Nishida, T. Yamagishi, Y. Nakamoto, Macromolecules 2010, 43, 3145. Crossref, CAS
21)	T. Ogoshi, Y. Nishida, T. Yamagishi, Y. Nakamoto, Macromolecules 2010, 43, 7068. Crossref, CAS
22)	T. Ogoshi, Y. Hasegawa, T. Aoki, Y. Ishimori, S. Inagi, T. Yamagishi, Macromolecules 2011, 44, 7639. Crossref, CAS
23)	P. C. Innis, F. Masdarolomoor, L. A. P. Kane-Maguire, R. J. Forster, T. E. Keyes, G. G. Wallace, J. Phys. Chem. B 2007, 111, 12738. Crossref, Medline, CAS
24)	T. Ogoshi, T. Aoki, S. Ueda, Y. Tamura, T. Yamagishi, Chem. Commun. 2014, 50, 6607. Crossref, Medline, CAS
25)	T. Ogoshi, T. Aoki, R. Shiga, R. Iizuka, S. Ueda, K. Demachi, D. Yamafuji, H. Kayama, T. Yamagishi, J. Am. Chem. Soc. 2012, 134, 20322. Crossref, Medline, CAS
26)	T. Ogoshi, K. Kitajima, T. Aoki, S. Fujinami, T. Yamagishi, T. Nakamoto, J. Org. Chem. 2010, 75, 3268. Crossref, Medline, CAS
27)	T. Ogoshi, H. Kayama, T. Aoki, T. Yamagishi, R. Ohashi, M. Mizuno, Polym. J. 2014, 46, 77. Crossref, CAS
28)	J. Chen, N. Li, Y. Gao, F. Sun, J. He, Y. Li, Soft Matter 2015, 11, 7835. Crossref, Medline, CAS
29)	T. Ogoshi, R. Sueto, M. Yagyu, R. Kojima, T. Kakuta, T. Yamagishi, K. Doitomi, A. K. Tummanapelli, H. Hirao, Y. Sakata, S. Akine, M. Mizuno, Nat. Commun. 2019, 10, 479. Crossref, Medline
30)	K. Kato, K. Onishi, K. Maeda, M. Yagyu, S. Fa, T. Ichikawa, M. Mizuno, T. Kakuta, T. Yamagishi, T. Ogoshi, Chem.—Eur. J. 2021, 27, 6435. Crossref, Medline, CAS
31)	T. Ogoshi, K. Demachi, K. Kitajima, T. Yamagishi, Chem. Commun. 2011, 47, 10290. Crossref, Medline, CAS
32)	C. Li, K. Han, J. Li, H. Zhang, J. Ma, X. Shu, Z. Chen, L. Weng, X. Jia, Org. Lett. 2012, 14, 42. Crossref, Medline, CAS
33)	W. Si, L. Chen, X.-B. Hu, G. Tang, Z. Chen, J.-L. Hou, Z.-T. Li, Angew. Chem., Int. Ed. 2011, 50, 12564. Crossref, Medline, CAS
34)	N. L. Strutt, H. Zhang, S. T. Schneebeli, J. F. Stoddart, Chem.—Eur. J. 2014, 20, 10996. Crossref, Medline, CAS
35)	Y. Ma, L. Chen, C. Li, K. Müllen, Chem. Commun. 2016, 52, 6662. Crossref, Medline, CAS
36)	T. Ogoshi, Y. Shimada, T. Akutsu, T. Yamagishi, Polymer 2017, 128, 325. Crossref, CAS
37)	W. Cui, H. Tang, L. Xu, L. Wang, H. Meier, D. Cao, Macromol. Rapid Commun. 2017, 38, 1700161. Crossref
38)	Z. Wang, T. Chen, G. Wang, W.-B. Hu, Y. A. Liu, H. Yang, K. Wen, ACS Appl. Polym. Mater. 2020, 2, 5566. Crossref, CAS

bradleypasioned87.blogspot.com

Source: https://www.journal.csj.jp/doi/10.1246/bcsj.20210243

Share this post