Self-Assembled Supramolecular Channels: Toward Biomimetic Materials for Directional Translocation

Self-Assembled SupramolecularChannels: Toward Biomimetic Materialsfor Directional Translocation

Yves-Marie Legrand[a] and Mihail Barboiu[a]*[a]Adaptive Supramolecular Nanosystems Group, Institut Europeen des Membranes,ENSCM-UMII-UMR CNRS 5635, Place Eugene Bataillon CC047, 34095, Montpellier(France)E-mail: [email protected]

Received: June 5, 2013published online: ■■

ABSTRACT: This Personal Account summarizes the recent developments in the developmentof self-assembled supramolecular channels and their dimensional extension towards up-scaledself-organized materials. This Personal Account begins with a short, non-exhaustive descriptionof artificial supramolecular channel systems that are involved in water-, proton-, and ion-transport processes through bilayer membranes. Then, these “all-made” artificial systems willbe described as a source of inspiration, by presenting several breakthroughs over the last fewyears in the field of biomimetic supramolecular channel systems. Their inclusion in artificialpolymeric/hybrid matrixes, which results in the formation of biomimetic artificial materials fordirectional translocation through channeling pathways, will be described in the last part of thePersonal Account, with an emphasis on all of the efforts that are necessary to maintain theirchannel-transporting function within bilayer membranes under up-scaled operating conditions.DOI 10.1002/tcr.201300011

Keywords: hybrid materials, ion channels, membranes, self-assembly, water channels

1. Introduction

Most of the vital bodily functions depend on the selectivetransport of metabolites between the cell and its exterior.[1] Thetransport of essential metabolites depends on the unique prop-erties of the hydrophilic inner domains of the protein channels,which play a crucial role in such translocation events overnanometer distances. These events are related to their complexinteractional mechanisms, although the solute-protein interac-tions at the molecular level are not yet fully understood.[2] Suchfeatures are illustrated by the functional complexity of self-organized membrane proteins, which may assist in proton, ion,and molecular translocation through membranes.[3] Artificialion channels, dimensionally and functionally, that fit the thick-

ness of the insulating bilayer of a membrane (30–35 Å) havebeen extensively studied with the hope of mimicking thenatural ionic conduction through protein channels.[4–18]

Within this context, functional supramolecular systems,which represent multiple copies of the specific recognition eventsbetween reversibly interacting molecular components, can generateadaptive translocationnetworkswith increaseddimensionalbehaviorthrough self-organization processes (Figure 1). This property affordsthe possibility of extending and engineering multiple supramole-cular interactions to generate and control the organization offunctional membrane materials across an extended scale. Moreover,it opens up wide perspectives for imagining interesting transitions

T H EC H E M I C A L

R E C O R D

Chem. Rec. 2013, ••, ••–•• © 2013 The Chemical Society of Japan and Wiley-VCH, Weinheim www.tcr.wiley-vch.de

from molecular-recognition functions to supramolecular self-assembly processes toward materials with directional translocationpathways,[7–9] which might have great potential in various applica-tions, such as membranes, sensors, information transfer devices, etc.

2. Artificial Supramolecular Channels

2.1. Self-Assembled Ion-Channels

Numerous artificial transport systems that utilize carriers,channel-forming or self-organized superstructures that are ableto orient, select, and pump the ion transport across membraneshave been developed over the last few decades.

The specific geometrical ion-recognition properties ofcrown ethers, as well as their propensity to form self-assembledfunctional architectures, made them good candidates forforming ion channels to be inserted within bilayer membranes.Many critical parameters may arise when it comes to ionsdiffusing across narrow hydrophobic pores. In a classical diffu-sion model, the ions diffuse through the membrane and themechanism is governed by the interaction equilibrium of thehydrated ions that lose part of the hydration shell in exchangefor the specific chelating groups that are located in the mem-brane. Fluorescence and amperometric techniques can be usedto quantify transport across the synthetic membranes, therebyproviding information on the selectivity, resistance, lifetime,and eventually the size and distribution of the channels. Aninteresting aspect of such channels is in the fact that it islockable by a permeant cation, which is reminiscent of iondeactivation in biological pores.

The pioneering work of Gokel’s group on syntheticamphiphilic crown ethers introduced the concept of“hydraphyles”, a class of macrocyclic ion channels that arewell-known as both in vitro[4] and in vivo[5,6] synthetic pores inphospholipid bilayers, in which ions are envisioned to diffusealong the crown-ether pathways (Figure 2a). Another seminalwork by Voyer and co-workers involved an interesting tubularstack of crown ethers that exhibited the formation ofamphipathic helical peptide carriers, which are expected to actas membrane-disrupting agents and display cytolytic activity(Figure 2b).[7,8]

By using similar strategies, several teams improved ourunderstanding of the transport properties, which allowedincreased metabolite pumping and enhanced transport selec-tivity. Synthetic ion channels and pores that were formed byrigid-rod molecules have been used by Matile and co-workersas key scaffolds for the synthesis of rigid-rod beta-barrel pores.The description of internal and external design strategies forthese rigid-rod beta-barrels covers a rich collection of pH-,pM-, voltage-, ligand-, and enzyme-gated synthetic multifunc-tional pores that can act as hosts, sensors, and catalysts(Figure 2c).[9,10] Ghadiri and co-workers developed cyclic d,l-α-peptides that acted preferentially on Gram-positive and/orGram-negative bacterial membranes compared to mammaliancells (Figure 2d).[11] Hybrid bio-assisted systems, such as thecrown-ether–Gramicidin A (gA) systems used by Koert’sgroup,[12] were found to be suitable structures for the design ofsuch synthetic channels, because they formed ion channelswith higher ion selectivity for K+ than for Cs+.

Barboiu and co-workers recently designed a versatileclass of ureido-crown ethers that were successfully usedas active transporters through liquid-,[13,14] bilayer,[15] orhybrid solid-membrane systems.[16] These transporters includeheteroditopic receptors that self-organize in solution andin the solid state, based on three encoded features: 1) The

Fig. 1. Up-scale transposition of self-organized functional ion channelsfrom supramolecular architectures toward informed materials for directionaltranslocation.

Mihail Barboiu received his PhD in1998 from the University of MontpellierII, before spending two years as an Asso-ciate Professor with Prof. Jean-MarieLehn at the College de France, Univer-sity Louis Pasteur, Strasbourg. Since2001, he has been a CNRS ResearchFellow, a Group Leader, and then aCNRS Senior Researcher at the InstitutEuropéen des Membranes in Montpel-lier, France. A major focus of his researchis Dynamic Constitutional Chemistry toward the develop-ment of Dynamic Interactive Systems for use in functionaladaptive biomimetic membranes and biosensors. Dr. Barboiuis the author of more than 180 scientific publications and hereceived the EURYI Award in Chemistry in 2004.

www.tcr.wiley-vch.de Chem. Rec. 2013, ••, ••–•• © 2013 The Chemical Society of Japan and Wiley-VCH, Weinheim

T H E C H E M I C A L R E C O R D

molecular-recognition sites for the anion and the cationare covalently linked (Figure 3a); 2) the guiding supramolecu-lar interaction is a urea head-to-tail H-bond association(Figure 3b); and 3) the urea allows the self-assembly of thesesystems within hydrophobic membranes and the selectivetransport of cations across the bilayers (Figure 3c). In thebilayer membranes, at low concentrations of the ureido-crownether, membrane disruption is observed, together with raresingle-channel openings, whereas, at higher concentration, arich array of interconverting channel conductance states isobserved. The channel formation is interpreted as arising fromdiscreet stacks of ureido-crown ethers, where the transport ofcations would occur through the macrocycles, admixed with

larger pores that are formed by association of the crown etherhead-groups around a central large pore.[15]

Another interesting example to be considered withinthis context is the G-quartet, that is, the hydrogen-bondedmacrocycle that is formed by the self-assembly of guanine/guanosine and stabilized by cations.[17] The G-quartet architec-ture represents a nice example of a system in which theG-quartet is dynamically exchanging with G-quadruplexes andthe linear ribbons (Figure 4a).[18] The role of cation templatingis to stabilize and amplify the G-quadruplex, the columnardevice that is formed by the vertical stacking of G-quartets,through coordination to the eight carbonyl oxygen atoms oftwo sandwiched G-quartets.

Fig. 2. Synthetic supramolecular channels within bilayer membranes: a) “Hydraphile” crown-ether channels thatfunction as both in vitro and in vivo synthetic ion pores. b) A tubular stack of crown-ethers that exhibits the formationof amphipathic peptidic ion-channel helices. c) Rigid-rod molecules that are used as key scaffolds for the synthesis offunctional beta-barrel pores. d) Tubular self-assembled cyclic peptides that form transmembrane ion channels, whichare reminiscent of cytoskeletal microtubules and viral coat proteins. Reproduced with permission from Refs [4–11].


S e l f - A s s e m b l e d S u p r a m o l e c u l a r C h a n n e l s

Fig. 3. Self-assembled ureido-crown-ethers a) for the simultaneous molecular recognition of anions and cations;b) for directional self-assembled superstructures; and c) for the selective transport of cations as ion channels acrossbilayers. Reproduced with permission from Refs [13–15].

Fig. 4. G-quadruplex as a dynamic architecture for single ion channels: a) G-quartets (6) dynamically exchange withG-quadruplexes and linear G-ribbons A and B. Reproduced with permission from Ref [18]. b) Active oligomers (9)that were obtained from functional self-assembled G-quadruplexes (8) of allyl-derived guanosine (7) and presentoptical activity within bilayers and conductance states across synthetic bilayer membranes. Reproduced with permis-sion from Ref [19].



Because they can accommodate ions at the center ofthe stacked tetrads, interesting ion-conductance behaviorsof the G-quadruplexes in the bilayer membranes have beenemphasized by Davis and co-workers only very recently(Figure 4b).[19] The synergistic use of metathesis synthetic pro-cedure and the inclusion of assembled guanosine oligomersin a hydrophobic bilayer environment have been proposedto obtain the ion-channel conductance of assembled G-quadruplexes in bilayers. These results show evidence of finitefunctional oligomers of G-quadruplexes that are embedded inliposomes by designing columnar transporting pathways forcations through membranes. Subsequently, architectures basedon this folate quadruplex have been developed to generatechannels across bilayers membranes.[20] Indeed, dendritic folaterosettes self-assemble, thereby giving rise to ionic channelsthat are studied in egg yolk phosphatidylcholine (EYPC) largevesicles.

2.2. Cation-π Interactions in Ion Channels

Intermolecular interactions that involve aromatic rings are keyprocesses in both chemical and biological recognition. Among

these interactions, cation-π interactions between positivelycharged species (alkali, ammonium, and metal ions) and aro-matic systems with delocalized π-electrons are now recognizedas important non-covalent binding forces of increasing rel-evance (Figure 5). Santarelli et al. developed an electrostaticmodel that describes trends in binding energy, based on differ-ences in electrostatic attraction. Interestingly, it shows thatinteraction energies of cation-π pairs correlate well with elec-trostatic potential above the π face of arenes (see the nicotine-tryptophan cation-π interactions (10) within the binding siteof acetylcholine receptors; Figure 5a).

This result correlates directly with transport in channelsbecause the inner facets of the pores are covered with functionalπ-moieties in direct contact with travelling cations. Extensivework by Dougherty et al. has given a clearer idea of the impor-tance of such subtle interactions, yet unequivocally critical.They showed the incorporation of unnatural amino acids intoa voltage-gated Na+ channel and demonstrated that a cation-πinteraction was responsible for the obligate nature of an aro-matic group at this position in the Na+ channels.[21]

Over the past 40 years, a large number of macrocyclicreceptors have been synthesized and evaluated for their ability

Fig. 5. Cation-π interactions in channels and in the solid state. a) Nicotine forms a cation-π interaction with atryptophan residue (10) in the binding site of acetylcholine receptors in the brain. C&EN, Copyright AmericanChemical Society, 2013. b) Chemical (11) and solid-state structures (12) of Gokel’s bis-tryptophan-crown-ether inthe absence (center) and presence of KI (right). Copyright Royal Society of Chemistry, 2003. c) Crystal structure ofthe salt complex, which shows clear cation-π interactions between the macrocyclic complexed K+ cation and the indolegroup of the phenylureidoindole (13). d) Two orthogonal views of the CPK representation of K+-indole cation-πinteractions for Barboiu’s (left) and Gokel’s complexes (right); K+ ions are represented as magenta spheres. CopyrightRoyal Society of Chemistry, 2003.



to transport cations and increased attention has been directedtowards macrocyclic receptors for anions. One strategy tocircumvent this tendency is to design heteroditopic receptormolecules with specific cation- and anion-binding sites thatcan simultaneously bind both of the salt ions[13,14,22] or to use[cation-carrier][anion-carrier] complexes as mixed carriers inmembrane transport experiments.[23]

During the last few decades, Gokel and co-workersmade crucial advancements and have provided useful insightsin this field.[24] Their studies were confined to macrocyclicequatorially bound alkali cations 11 and 12 (Figure 5b,d).Several groups, including our own,[25–27] have found that theindole is a strong π-donor group for the K+ ion. Barboiu andco-workers showed that, in terms of the dynamic diversity,the heteroduplex complex of the macrocycle-cation-π and itsanion-ureidoarene counteranion (13) represented an attrac-tive illustration of the self-selection, based on the specificcation-indole and not cation-benzene π-interactions, of aunique solid-state component. Interestingly, without anyconstraining steric impediment, the pyrolo subunit of theindole (i.e., the C2=C3 bond) is a strong π-donor group, asdemonstrated previously (Figure 5c,d).[23] This determinesimportant differences between cation transport performanceswhich can be attributed to the stability of the resulted[cation-carrier][anion-carrier] complexes in the membranephase. Among all of the anion carriers, the highest observedselectivity is related to the indole-type (tryptophan-derived)compound, as previously demonstrated by the structuralstudies of Gokel and Mukhopadhyay[25] and Barboiu andco-workers.[27]

2.3. Water and Proton Channels

Water is fundamental to life and performs a variety of func-tions related to its complex dynamic behavior when waterclusters interact with biomolecules.[2] The combination ofnatural proteins that present high water-conductance stateswithin natural conditions with artificial lipidic or polymericmatrices has demonstrated that natural aquaporins can beused as bio-assisting building blocks for the construction ofhighly selective artificial membranes.[28] Until now, very fewexamples of artificial water channels that integrate only syn-thetic elements in their water-selective translocation unit havebeen reported.[29]

Synthetic building blocks have been used to generatesuch systems, in which water was very efficiently transportedthrough hydrophilic, hydrophobic, or hybrid hydrophobic/hydrophilic tubular superstructures that were insertedwithin the bilayer membranes. Dendritic dipeptides (14)self-assemble through enhanced peripheral π-stacking toform stable cylindrical helical pores (14.5 Å in diameter)(Figure 6a). These pores selectively transport water molecules

against ions through self-assembled hydrophobic nanotubesthat are stable in phospholipid membranes. The ion-exclusionphenomena are based on hydrophobic effects, which appearto be very important.[30] Later, Barboiu and co-workersreported that imidazole (I) quartets (15) can be mutually sta-bilized by inner dipolar water wires (Figure 6b).[31] TheI-quartets are stable in solution, in the solid state, and withinbilayers, thus leading to tubular channel-type chiral super-structures. These systems have provided excellent reasons toconsider that the supramolecular chirality of I-quartets andwater-induced polarization within the channels may bestrongly associated. Then, Hou and co-workers proposed avery elegant artificial system that functions exclusively assingle-molecular water channels (Figure 6c).[32] Hydrazide-incorporated pillar[5]arenes (16) were used to form H-bonded superstructures that are robust when embedded inbilayer membranes.

Unlike ions and water molecules that can be transportedacross membranes as isolated entities, protons need a carrieron which the charged monoatomic particle hops from one toanother. Water can be the appropriate media for such trans-port (in its hydronium form), as can many other organicmoieties. Although several well-known membrane exampleshave been identified to transport protons, such as sulfonic-acid functions in Nafion or proton pumps in transmembraneATPases, rare examples of synthetic bilayer membranes thattransport H+ ions can be found. Hou’s group has developedartificial transmembrane channels based on the same pillar-arene superstructures in which water forms wires. It wasshown that proton migration occurred along the water wireaccording to the Grotthus mechanism[33] by using the isotopeeffect of H versus D (deuterium). In Barboiu’s team, ureidoimidazoles have been used to construct I-quartets that aremutually stabilized by inner water dipolar wires in a similarmanner to the way that alkali cations stabilize the G-quadruplexes.[34] The H-bonding of these I-quartets leads totubular solid-state structures and, in a membrane environ-ment, to a barreled channel. Within the I-quartet nanotubes,water molecules of unique dipolar orientation can preservethe electrochemical potential along the channel. These resultsindicate that protons can permeate bilayer membranesthrough I-quartet channels. No example of proton transportacross a bilayer membrane with a different media than water-wires could be found.

As shown above, specific recognition events betweenvarious molecular components can generate a wide range ofreversible interactions within self-assembled architectures,which, in turn, might induce adaptive supramolecular net-works for translocation. Next, we will focus our attention onthe generation and control of the organization of functionaltransporting materials across an extended scale or of dynamicinteractive systems.[35]



3. Biomimetic Artificial Materials forDirectional Translocation

3.1. Hybrid Materials for Molecular Recognition

By using efficient systems that were tested on the molecularlevel, as described above, a higher level of complexity can beobtained through the combination of multiscale strategies.Although understanding molecular natural pumps was amaz-ing, imitating them is highly challenging and the questionremains can we really achieve the integration of such isolatedelements in functional materials? One general strategy that isadopted by Barboiu and co-workers is to adapt the basics ofconstitutional chemistry for the construction of hierarchicalmembrane channel systems over large (μm) scales (Figure 1).Rather than aiming for the most-rigid and robust materials,it is accepted that some natural processes do a very good jobat constructing molecular assemblies, all the way to highlyordered macroscopic objects (e.g., single crystals). Based on

this postulate, a great deal of molecular design is necessary tointegrate the right functionalities at the right place. The natureof the components can be readily varied because the stablepolymeric backbones can be decorated with an immensevariety of organic functions. The hydrogen-bonding interac-tions between organic molecules and the silica precursor aredeterminant for the morphology of the resulted superstructuresin the sol–gel transcription processes by using heterogeneousorganogelator-TEOS systems.[36,37,38]

The chemistry of such hybrid membrane transportsystems that are of interest for molecular information transferhas been extensively developed during the last twenty years.The membrane selectivity may either be induced by carriermolecules or by transmembrane channels. In the importantpapers concerning the transport mechanism in membranes,the concept of carrier in mostly associated with liquid mem-branes and the concept of channels with bilayers membranes.From a mechanistic point of view, we may use carriers that

Fig. 6. Examples of water-conducting pores: a) Helical pore, which is assembled from a dendritic peptide. Repro-duced with permission from Ref. [30]. b) I-quartets, which present supramolecular chirality and accommodate dipolarwater wires along the length of the channel. Reproduced with permission from Ref. [31]. c) The hydrazide units formpentameric cylinders. Reproduced with permission from Ref. [32].



self-assemble into functional aggregates, which would presentcombined (hybrid) intermediate features between the formercarrier-monomers and the resultant pseudo-channel-formingstructures. As part of the program concerned the description ofefficient membrane systems that affected the selective transportof biological relevant species, we became interested in fixed-sitehybrid membranes. By using this strategy, multiple molecularrecognition in solution could be transferred into solid densethin-layer hybrids. The field of hybrid supramolecular mem-branes covers all materials whose silsesquioxane precursorsare initially connected through weak reversible non-covalentinteractions so that they spontaneously undergo dynamic self-assembly and disassembly processes. Grafted polytopic fixed-site organic complexants could improve the selectivity andthe solute flux across the hybrid membrane, owing to a more-specific complexation, by combining different non-covalentinteractions. Hybrid heteropolysiloxane materials are of inter-est for the preparation of supramolecular solid dense materialsat the molecular level. These materials are based on theincorporation of suitable molecular components for multiplerecognition transport-based functions. In these fixed-sitecomplexant membranes, the fixed receptor is not a carrier;rather, it just selectively assists the solute diffusion (D) in themembrane by selective complexation-decomplexation reac-tions (CDR) at the receptor-site level. Polytopic macrocyclicreceptors are able to facilitate the transport of metal ionsthrough specific selective interactions.[39] The multiple molecu-lar recognition of amino acids[40] or bioorganic acids[41] canbe transferred into dense solid materials by anchoring thepolytopic macrocycles in the siloxane matrix through sol-gelprocesses. Of particular interest is the potential ability of suchthin-layer solid films to show functional molecular recognitionproperties and the simple generation of the directional conduc-tion pathways through self-assembly so as to enable efficienttranslocation events. This premise hold true for suitablydesigned thin-layer hybrid membranes, thereby minimizingthe distance between the receptors by self-organization, withthe hope of reducing non-selective solute-diffusion (D) andcreating functional favorable diffusion patterns for the mobilesolute into the solid dense material.

3.2. Directional Ion Pathways in NanoscopicHybrid Materials

One interesting objective is the development of straightforwardstrategies for maintaining and expanding the self-assemblyfunctional behavior of related ion-channel-based systems thatpresent a conductance state in bilayer membranes toward direc-tional nanoscopic pathways in functional materials. The firstapproach towards this objective is to control the synergeticsupramolecular self-organization of molecular building blockswith further covalent sol-gel modification, which might be

considered as a useful approach for the design of functionalself-organized hybrid materials. By using silsesquioxane precur-sors 17 (Figure 7), in which the functional organic and siloxanemoieties are covalently connected, during the sol-gel process,the weak supramolecular interactions of the molecular compo-nents are typically less robust and they usually are destabilizedwhen the cross-linked covalent siloxane network forms. Thehierarchical dynamic self-assembly of crown-ether ribbon-type architectures in solution/gel leads, in a second sol-geltranscription/resolution step, to lamellar solid hybrid materialson the nanoscopic scale.[16]

These dynamic crown-ether ion channels (17) can be“frozen” in self-organized hybrid matrixes by using sol-gelprocesses, thereby resulting in the formation of solid densehybrid membranes. The self-organized nanoscale channels inthe hybrids (18) define a particularly attractive functionaltransport device that encodes the required information forboth directional-diffusion cation- (tubular macrocycles) andanion-transport (sandwich-urea) mechanisms in the hybridmembrane material (Figure 7). It results in the formation ofnanoscale translocation pathways, similar to the way that pore-type proteins assist ion diffusion along the cell membrane.These membranes have been used to transport adenosine tri-phosphate (ATP2–) anions upon a synthetic ion-driven pump,“fuelled” by a Na+ concentration gradient.

On the other hand, the weak non-covalent interactions(H-bonds, van der Waals, etc.) that control the positioning ofmolecular components in supramolecular networks are typi-cally less robust than the cross-linked silica network. Accord-ingly, a sole solution is to improve the binding (association)efficiency because, at least in theory, an increased numberof interactions might improve the stability of self-assembledstructures that communicate with the inorganic siloxanenetworks during the formation of the silica matrix. Amongthese systems, nucleobases and nucleosides are well-known,fascinating compounds that generate multiple complementaryH-bonding interactions, as well as hydrophobic and stackinginteractions. Homo- and heteropairing of adenine and uracilderivatives result in the formation of interconverting librariesof dimers, trimers, and oligomers through the combination ofH-bond pairs. The resolution of such dynamic supramolecularlibraries may be achieved by sol-gel processes, thereby resultingin the amplification of the most-compact architectures, owingto the combined hydrophobic/H-bonding affinities in the finalconstitutional structure of the resultant hybrid materials, com-pared with the unpolymerized powders.[42] This strategy hasalso been used to transcribe the supramolecular chirality of theG-quadruplex (20) on the nanoscale and microscale to obtainchiral hybrid materials (21) by using achiral guanosine-heteropolysiloxane starting components (Figure 8).[43,44]

Stacked G-quartet superstructures that are stabilized bytemplating K+ cations induce an important increase in the



conductivity by a factor of 1000 of G-quartet membrane filmscompared to the non-templated G-ribbon membrane, whichpresents very low conductivity.[45] Base stacking, accompaniedby π-π interactions, may induce coherent charge mobility[46] orpolarizability[47] in stable G-quartet materials.

The structural functionality of channel proteins is notdefined by the complex receptor binding sites. Very simplefunctional moieties (carbonyl, hydroxy, amide, etc.) are ori-ented toward the protein core, surrounding the transportdirection. We recently showed that simple molecules 23–27(Figure 9), which collectively define transport devices by self-assembly, could be successfully used to transfer the overallfunctionality of their supramolecular self-organization inhybrid membrane materials by sol-gel transcription. We haveshown that specific H-bonding communication between these“peptoid hybrid networks” (Figure 9) resulted from the self-assembly of amino-acid-based herepolysiloxanes 23–27, whichhave the ability to create hydrophilic pathways of differentchemical properties in the self-organized superstructures ofhybrid membrane materials (MB26). These structures areessential for the diffusion process and for the selectivity of thetransport of hydrated alkali cations. Although these pathwaysdo not merge to cross the micrometric films, they are well-defined on the nanoscale.[48]

Proton transport is both a necessary step for vital functionsin biological environments and a highly promising process forindustry, for example in fuel cells. Proton conduction throughpolymeric membranes is well-established, but higher perfor-mances are targeted in this competitive area.[49] Interestingapplicative results have been obtained for proton-exchangemembrane (PEM) systems, in which the self-organizedhybrid sulfonate heteroplysiloxanes generates directionalproton-channel superstructures in a scaffolding hydrophobichybrid material.[50]

It has been shown that the controlled generation of con-nected self-organized channels of hundreds of nanometers inlength for directional proton diffusion represents a straightfor-ward approach for the design of a new class of PEM hybridmaterials with high ionic conductivities (equal or superior tothe reference Nafion 117 commercial membrane) whilst pre-senting good stability and high selectivity for the transport ofprotons over MeOH (fuel-cell solvent).[49] The simple synthesisprocedure, high proton/MeOH selectivity, as well as promisingproton conductivities of these hybrid membranes suggestthey may be promising candidates for use in Direct MethanolFuel Cells-DMFCs; however, they should be formed intomembrane electrode assemblies and tested in fuel cells toconfirm their suitability (Figure 10).

Fig. 7. Macrocyclic self-organized hybrid membrane materials: a) Self-organization in solution and the sol-geltranscription of encoded channels of heteropolysiloxane crown-ether ribbons (17) into a hybrid material (18).b) Cross-section micrograph of a hybrid membrane; the thin-layer dense film is deposited onto polymeric supports.TEM image of the surface and suggested model for the arrangement of the hybrid material; parallel disposition in theX-ray crystal packing of the macrocyclic receptors for both cation- (violet spheres) and anion-assisted diffusion (greenspheres) within the hybrid channels. Reproduced with permission from Ref. [16].



Fig. 8. Constitutional strategy to transcribe and fix the self-assembly of G-quadruplex architectures into self-organized nanohybrids. In the first resolution step, the G-quartet is pre-amplified in solution in the presence of metalcations from a dynamic pool of ribbon-type or cyclic supramolecular architectures. Then, in a second selection sol-gelstep, the G-quadruplex is irreversibly fixed in a siloxane inorganic network within a hybrid G-quartet material (21).Reproduced with permission from Ref. [44].

Fig. 9. “Peptoid” hybrid materials: a) Amino-acid-based heteropolysiloxanes 23–27 for the synthesis of “peptoidhybrid networks”. b) TEM images of hybrid material MB26. c) Crystal packing of compound 26 with a soft surface.Reproduced with permission from Ref. [48].



Fig. 10. Highly self-organized proton-exchange hybrid membrane/PEM systems: a) Structure and self-organizationof molecular precursors in hybrid functional PEM materials (34): a) Synthesis of precursors 30 and 33 and schematicrepresentation of their hierarchical self-organization in solution, followed by the sol-gel transcription of encodedmolecular features into hybrid membrane materials. b) Image of mechanically stable homogeneous membranes andc) Proton conductivity at 25°C, with 100% relative humidity of the hybrid self-organized membranes. Reproducedwith permission from Ref. [50].



3.3. Dynamic Constitutional Hybrids

Dynamic constitutional hybrids are materials in which organic(supramolecular) and inorganic domains are reversibly con-nected through dynamic or non-covalent interactions.[36–38,51]

One strategy concerns the nanoscale confinement of organiccomponents within a preformed scaffolding inorganic matrix.Mesoporous materials are suitable for the development ofproperties related to confinement or anisotropic orientation(Figure 11).[36]

By using the channel-forming macrocyclic systems describedabove, but confined within a solid mesoporous membrane matrix,

remarkable results in terms of selectivity were obtained and, mostimportantly, theadaptabilityof thesematerials towards their environ-ment and, more specifically, towards the solution to be filtered wasconfirmed.[36] Indeed, the constitutionally confined columnar ion-channel architectures that are reversibly confined within scaffoldinghydrophobic silica mesopores can be structurally characterized byusing X-ray diffraction and morphologically tuned by templatingalkali salts. The dynamic character, which is due to reversible inter-actions between the continually interchanging components, causesthematerials torespondtoexternal ionicstimuliandtoadjust to formthe most-efficient transporting superstructure in the presence of the“fittest” cation, as selected from a set of diverse less-selective possiblearchitectures that can form through self-assembly.

Evidence has been presented that such a membraneadapts and evolves its internal structure so as to improveits ion-transport properties: The dynamic non-covalentlybonded macrocyclic ion-channel-type architectures can bemorphologically tuned by templating alkali salts during thetransport experiments or the conditioning steps. From a con-ceptual point of view, these membranes express a synergisticadaptive behavior: The simultaneous binding of the fittestcation and its anion would be a case of “homotropic allostericinteractions” because, in time, it increases the transport effi-ciency of the pore-contained superstructures through a selec-tive evolution process toward the fittest ion channel.

This above-proposed strategy has also been applied to theconstruction of thin-layer mesoporous electrodes.[52] Non-covalent confined fullerenes in scaffolding inorganic mesoporesprovide a very useful device that is able to dynamically performconstitutional conduction functions. It is worth mentioningthat the ability of fullerene wires to quickly uptake/slowlyrelease electrons is based on the redox behaviors of the C60components that are in close contact under the confinementconditions. We note that dynamically confined fullerene wiresshow a capacitive diffusion of electrons during the transporttime. Note that these nanoscale systems might exhibit a dispo-sition toward the functional tunability of the surface or com-position of the pore that can be altered in order to control theflux or the gating functions.

Another strategy for preparing dynamic constitutionalhybrids concerns the use of a reversible non-covalent interfacebetween the supramolecular/organic and siloxane/inorganicconstituents during the sol-gel process. Because the self-organization of organic/inorganic domains has to explore thehypersurface of all of the structure/energy combinations, thebuild-up of the final hybrid materials through a growth/self-reparation/termination sequence might be able to select thecorrect spatial geometries for the supramolecular organic andinorganic entities that emerge from a collection of buildingblocks. Similar “communication processes” have been identifiedin DNA transcription into inorganic materials.[53] A dynamicreversible interface might mediate the structural self-correlation

Fig. 11. Dynamic constitutional hybrid membranes: Generation of direc-tional ion-conduction pathways that can be morphologically tuned bytemplating with alkali salts within dynamic hybrid materials that are obtainedby the hydrophobic confinement of ureido-macrocyclic receptor 35 withinsilica mesopores. Anodisc alumina membranes (AAM) can be filled with asurfactant and a silica precursor to provide hexagonal ordered networks andsilica mesopores, respectively. After heat treatment, only highly ordered silicamesopores are left in the macropore of the AAM commercial membrane. Thesilica mesopores can be functionalized, for example, with hydrophobicoctadecyltrichlorosilane (ODS), which provides a similar environment to theinterior of phospholipid membranes, the main difference being the thicknessof the membrane (100.000-times ticker). Reproduced with permission fromRef. [36].



of supramolecular and inorganic domains by virtue of theirbasic constitutional behaviors. The resultant hybrid materialmight undergo continual change in its constitution, throughthe dissociation/reconstitution of different mesophases duringthe sol-gel process. Such constitutional heteroditopic ureido-crown-ether hybrid channel (18) and G-quartet (21) materialshave been prepared under thermodynamic control, whichcontain supramolecular ion-channel-type architectures, com-bined with hydrophobic siloxanes as precursors for the inorganicsilica matrix (Figure 12).[54]

4. Conclusion

In conclusion, herein, we have discussed the latest strategiesfor transcribing and fixing supramolecular ion-channel-typearchitectures in self-organized and constitutional hybridmaterials. In particular, the communication of covalent orreversible interfaces between organic/supramolecular andinorganic/siloxane networks represents a useful strategy forimproving their compatibility and self-organization over

enhanced dimensions. After the sol-gel process, interpen-etrated hybrid components lead to biomimetic membranematerials in which the features of the supramolecular channeland the inorganic networks are expressed through cross-overand linear-processing schemes. The dynamic self-assembly ofsupramolecular channel systems that are prepared under ther-modynamic control may, in principle, be connected to kineti-cally controlled sol-gel or dynamic covalent polymerizationprocesses to perform selective transport functions. Such a“dynamic marriage” between supramolecular self-assemblyand inorganic sol-gel polymerization processes that synergis-tically communicate leads to higher self-organized hybridchannel materials with increased micrometric scales. Moregenerally, applying such considerations to hybrid materialsleads to the definiotion of constitutional hybrid materials, inwhich organic (supramolecular)/inorganic domains reversiblycommunicate over large distances. This report might providenew insight into the basic features that control the design offunctional constitutional hybrid materials for directionaltranslocaton.[55,56] Considering the simplicity of this strategy,

Fig. 12. Sol-gel process of a) classical and b) constitutional hybrid materials, based on (i) G-quartet (21) and (ii)ureido-crown-ether (18) architectures. Reproduced with permission from Ref. [54].



possible applications in the synthesis of more-complex hybridarchitectures might be very effective, thereby getting closer tonew expressions of complex bioinspired matter.

Acknowledgement

This work was supported by funds from the ITN MarieCurie network–DYNANO (PITN-GA-2011-289033, http://www.dynano.eu) and by the ANR (2011 BS08 006 04-MULTISELF project).

REFERENCES

[1] P. Ball, Chem. Rev. 2008, 108, 74–108.[2] D. Zhong, S. K. Pal, A. H. Zewail, Chem Phys. Lett. 2011, 503,

1–11.[3] S. Cukierman, Biophys. J. 2000, 78, 1825–1834.[4] A. Nakano, Q. Xie, J. V. Mallen, L. Echegoyen, G. W. Gokel,

J. Am. Chem. Soc. 1990, 112, 1287–1289.[5] S. Negin, B. A. Smith, A. Unger, W. M. Leevy, G. W. Gokel,

Int. J. Biomed. Imag. 2013, Article ID 803579.[6] G. W. Gokel, I. A. Carasel, Chem. Soc. Rev. 2007, 36, 378.[7] Y. Vandenburg, B. D. Smith, E. Biron, N. Voyer,

Chem.Commun. 2002, 1694–1695.[8] P.-L. Boudreault, N. Voyer, Org. Biomol. Chem. 2007, 5,

1459–1465.[9] S. Matile, Chem. Soc. Rev. 2001, 30, 158–167.

[10] N. Sakai, J. Mareda, S. Matile, Acc. Chem. Res. 2005, 38, 79–87.

[11] D. T. Bong, T. D. Clark, J. R. Granja, M. R. Ghadiri, Angew.Chem. Int. Ed. Engl. 2001, 40, 988–1011.

[12] J. R. Pfeifer, P. Reiss, U. Koert, Angew. Chem. Int. Ed. 2006,45, 501–504.

[13] M. Barboiu, G. Vaughan, A. van der Lee, Org. Lett. 2003, 5,3073–3076.

[14] M. Barboiu, J. Incl. Phenom. Macrocycl. Chem. 2004, 49, 133–137.

[15] A. Cazacu, C. Tong, A. van der Lee, T. M. Fyles, M. Barboiu,J. Am. Chem. Soc. 2006, 128, 9541–9548.

[16] M. Barboiu, S. Cerneaux, A. van der Lee, G. Vaughan, J. Am.Chem. Soc. 2004, 126, 3545–3550.

[17] J. T. Davis, Angew. Chem. Int. Ed. 2004, 43, 668–698.[18] S. Pieracini, S. Masiero, O. Pandoli, P. Samori, G. P. Spada, G.

Gotarelli, Org. Lett. 2006, 8, 3125–3128.[19] L. Ma, M. Melegari, M. Colombini, J. T. Davis, J. Am. Chem.

Soc. 2008, 130, 2938–2939.[20] N. Sakai, J. Matile, J. Am. Chem. Soc. 2006, 128, 2218–2219.[21] V. P. Santarelli, J. Dougherty, J. Biol. Chem. 2007, 282, 8044–

8051.[22] M. Barboiu, N. D. Hovnanian, C. Luca, L. Cot, Tetrahedron

1999, 55, 9221–9229.[23] C. Arnal-Herault, M. Michau, M. Barboiu, J. Membrane Sci.

2008, 321, 94–99.[24] G. W. Gokel Chem. Commun. 2003, 2847–2852.

[25] G. W. Gokel, A. Mukhopadhyay, Chem. Soc. Rev. 2001, 30,274–286.

[26] O. M. Cabarcos C. J. Weinheimer, J. M. Lisy, J. Chem. Phys.1999, 110, 8429–8436.

[27] C. Arnal-Herault, M. Barboiu, E. Petit, M. Michau, A. van derLee, New J. Chem. 2005, 29, 1535–1539.

[28] M. Barboiu, A. Gilles, Acc. Chem. Res. 2013, 46, DOI:10.1021/ar400025e.

[29] M. Barboiu, Angew. Chem. Int. Ed. 2012, 51, 11674–11676.[30] M. S. Kaucher, M.Peterca, A.E. Dulcey, A. J. Kim, S.A.

Vinogradov, D. Hammer, P. A. Heiney, V. Percec, J. Am.Chem. Soc. 2007, 129, 11698–11699.

[31] Y. Le Duc, M. Michau, A. Gilles, V. Gence, Y.-M. Legrand, A.van der Lee, S. Tingry, M. Barboiu, Angew. Chem. Int. Ed.2011, 50, 11366–11372.

[32] X. B. Hu, Z. Chen, G. Tang, J. L. Hou, Z. T. Li, J. Am. Chem.Soc. 2012, 134, 8384–8387.

[33] B. L. de Groot, H. Grubmueller, Science 2001, 294, 2353–2357.

[34] J. T. Davis, G. P. Spada, Chem. Soc. Rev. 2007, 36, 296–313.[35] M. Barboiu, Chem. Commun. 2010, 46, 7466–7476.[36] A. Cazacu, Y. M. Legrand, A. Pasc, G. Nasr, A. van der Lee, M.

Barboiu, Proc. Natl. Acad. Sci. USA 2009, 106, 8117–8122.[37] S. Mihai, A. Cazacu, C. Arnal-Herault, G. Nasr, A. Meffre, A.

van der Lee, M. Barboiu, New. J. Chem. 2009, 33, 2335–2343.[38] M. Barboiu, A. Cazacu, S. Mihai, Y.-M. Legrand, G. Nasr, Y.

Le Duc, E. Petit, A. van der Lee, Microp. Mesop. Mat. 2011,140, 51–57.

[39] M. Barboiu, C. Luca, C.Guizard, N. Hovnanaian, L. Cot, G.Popescu, J. Membrane Sci. 1997, 129, 197–207.

[40] M. Barboiu, C. Guizard, N. Hovnanian, J. Palmeri, C. Reibel,C. Luca, L. Cot, J. Membrane Sci. 2000, 172, 91–103.

[41] M. Barboiu, C. Guizard, C. Luca, N. Hovnanian, J. Palmeri,L. Cot, J. Membrane Sci. 2000, 174, 277–286.

[42] C. Arnal-Hérault, M. Barboiu, A. Pasc, M. Michau, P. Perriat,A. van der Lee, Chem. Eur. J. 2007, 3, 6792–6800.

[43] C. Arnal-Herault, M. Michau, A. Pasc-Banu, M. Barboiu,Angew. Chem. Int. Ed. 2007, 46, 4268–4272.

[44] S. Mihai, Y. Le Duc, D. Cot, M. Barboiu J. Mater. Chem.2010, 20, 9443–9448.

[45] C. Arnal-Herault, A. Pasc-Banu, M. Michau, D. Cot, E. Petit,M. Barboiu, Angew. Chem. Int. Ed. 2007, 46, 8409–8413.

[46] A. Calzolari, R. Di Felice, E. Molinari, Solid State Commun.2004, 131, 557–567.

[47] H. Cohen, T. Sapir, N. Borovok, T. Molotosky, R. Di Felice,A. B. Kotylar, D. Porath, Nano Lett. 2007, 7, 981–986.

[48] M. Michau, M. Barboiu, R. Caraballo, C. Arnal-Hérault, P.Perriat, A. Banu, A. van der Lee, Chem. Eur. J. 2008, 14,1776–1783.

[49] M. A. Hickner, H. Ghassemi, Y. S. Kim, B. R. Einsla, J. E.McGrath, Chem. Rev. 2004, 104, 4587–4612.

[50] M. Michau, M. Barboiu, J. Mater. Chem. 2009, 19, 6124–6131.

[51] M. Barboiu, A. Cazacu, M. Michau, R. Caraballo, C.Arnal-Herault, A. Pasc-Banu, Chem. Eng. Process. 2008, 47,1044–1052.



[52] Y. Le Duc, A. Gilles, S. Mihai, V. Rouessac, S. Tingry, M.Barboiu, Chem. Commun. 2013, 49, 3667–3669.

[53] M. Numata, K. Sugiyasu, T. Hasegawa, S. Shinkai, Angew.Chem. 2004, 116, 3341–3345; Angew. Chem. Int. Ed. 2004,43, 3279–3283.

[54] S. Mihai, A. Cazacu, C.Arnal-Herault, G. Nasr, A. Meffre, A.van der Lee, M. Barboiu, New. J. Chem. 2009, 33, 2335–2343.

[55] Topics in Current Chemistry, Constitutional Dynamic Chemistry,Vol. 322 (Ed. M. Barboiu), Springer, Berlin, 2012.

[56] a) M. Barboiu, J. M. Lehn, Proc. Natl. Acad. Sci. U.S.A. 2002,99, 5201–5206; b) M. Barboiu, F. Dumitru, Y.-M. Legrand,E. Petit, A. van der Lee, Chem. Commun. 2009, 2192–2194;c) F. Dumitru, Y. M. Legrand, A. van der Lee, M. Barboiu,Chem. Commun. 2009, 2667–2669.



Documents

Self-Assembled Supramolecular Channels: Toward Biomimetic Materials for Directional Translocation