Molecular Encapsulation & Graphite|Multi Purpose “Magic Sand”

REVIEWS

Molecular Encapsulation
Fraser Hof, Stephen L. Craig, Colin Nuckolls, and Julius Rebek, Jr.*

Dedicated to Ivar Uginents. This review is about molecular aggregates of a certain sort, namely, those that assemble and more or less completely surround other molecules. Taking part in this intimacy imparts unique properties to the participants, and new functions emerge from the aggregate as a whole. For the most part, we emphasize self-complementa ry structures. Their ability to assemble–an expression of the molecule×s desire to be something more than it is–results from instructions engi- neered into the molecules during their creation.

To display how Encapsulation occurs

 

Louis Kahn, architect of the Salk Institute in La Jolla, said[1] TMeven a common, ordinary brick wants to be something more than it is.∫ Suppose that were also true of molecules. We know that they can and do aggregate; they give complex structures, and by doing so they acquire new properties– functions that may not be apparent from a study of the individual compo-

1. Introduction

In the 1980s, most of the publications on molecular recognition dealt with the selectivity of synthetic receptors and the energetics of intermolecular forces, and were confined for the most part to bimolecular systems. Termolecular systems would show up in a desultory way, such as in models for allosteric effects,[2] but there was much less work in pursuit of, for example, binding cooperativity.[3] There were rare cases in which a third molecule would interact with a weakly held bimolecular complex, and their beauty was exhilarating–like hitting a moving target.[4]

Studies of termolecular systems spread quickly during the 1990s in the form of template effects.[5±7] Interest in molecular self-replication had been ignited by von Kiedrowski[8] using nucleic acid components and the fire leapt to modified nucleic acids[9, 10] and entirely synthetic systems.[11±14] Even peptides now fan these flames.[15, 16] Elsewhere, other bimolecular

[*]Prof.J.Rebek,Jr.,F.Hof
The Skaggs Institute for Chemical Biology
The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, CA 92037 (USA) Fax: (1)858-784-2876
E-mail: jrebek@scripps.edu

S. L. Craig
Department of Chemistry, Duke University Durham, NC (USA)

C. Nuckolls
Department of Chemistry, Columbia University New York, NY (USA)

reaction templates were devised[17] and more complex systems were contrived. This is engineering (or is it art ? )[18] at the molecular level.[19] It did not matter that these systems weren×t particularly efficient, what mattered was that they improved the understanding of three-component systems.

Synthetic receptors became more sophisticated and con- cave surfaces such as clefts,[20] armatures,[21] tweezers,[22, 23] bowls,[24] and other shapes[25] emerged for the study of reversible interactions. We thought that a receptor could be created that could completely surround the target by using only the weak intermolecular forces of molecular recognition. These systems would be capsular assemblies, the reversible counterparts of the carcerands and cryptophanes–the cova- lently bound TMmolecules within molecules∫ crafted by Cram et al.[26] and Collet and co-workers.[27] The synthetic economy of using aggregates of self-complementary compounds[28±30][31] rather than one large molecule as a receptor proved irresistible. These structures have been termed TMencapsula- tion complexes∫ and they are now tools of physical organic chemistry on the nanoscale.

At the outset, most of the assemblies–spectacular as they were in the number of components, the intricacies, and sheer molecular weight–did little more than fill space. More sophisticated properties quickly emerged, and these will be the focus of this review. The complexes are used today as probes of isolated molecules and of the intrinsic character- istics of the liquid state, and are capable of enantioselective recognition, reversible polymerization, isolation of reactive species, and promoting reactions within their interiors.

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508 1 WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002 1433-7851/02/4109-1489 $ 20.00+.50/0 1489

REVIEWS

J. Rebek, Jr. et al.

Solvent-free environments provide an undiluted, often intensified interaction between the host and guest or between multiple guests themselves. Groups of molecules residing in the channels of zeolites, the pores of polymers, active sites of enzymes, and within globular micelles or dendrimers display behavior that is amplified by their environment. Encapsula- tion is a means by which the environment and encounters of a single molecule can be rigorously controlled. What can be learned about them in their isolated states, either alone or, as we shall relate, grouped in pairs? Much effort has been expended on isolation tactics in the chemical sciences. In the solid state or glassy states, inert matrices are used at low temperatures to isolate and stabilize reactive intermediates. In the gas phase, isolation can be achieved by subjecting a

J. Rebek, Jr. F. Hof

molecule to such low pressure that collisions with other molecules are essentially zero.[32] In the encapsulation com- plexes presented in this review, molecules are isolated from solvent encounters at ambient temperatures in the liquid phase.

Other supramolecular structures are also capable of surrounding guest molecules. The differences here are those of topology and timing. Supramolecular rings, tubes, and cavitands are able to briefly bind one or more guest molecules within a restricted environment, while still allowing varying amounts of solvent access to the secluded guests. We will not dwell on these here; rather, this review focuses on hosts that self-assemble and encapsulate molecules within a closed-shell topology.

S. Craig C. Nuckolls

Julius Rebek, Jr. was born in Hungary in 1944 and lived in Austria from 1945 ± 1949. He and his family then settled in the USA in Kansas. He received his undergraduate education at the University of Kansas in 1966, and obtained his Ph.D. degree from the Massachusetts Institute of Technology (1970) for studies in peptide chemistry with Professor D. S. Kemp. As an Assistant Professor at the University of California at Los Angeles (1970 ± 1976) he developed the TMthree-phase test∫ for reactive intermediates. In 1976 he moved to the University of Pittsburgh where he rose to the rank of Professor of Chemistry and developed cleftlike structures for studies in molecular recognition. In 1989 he returned to the Massachusetts Institute of Technology, where he was the Camille Dreyfus Professor of Chemistry, and devised synthetic, self-replicating molecules. In July of 1996, he moved his research group to The Scripps Research Institute to become the Director of The Skaggs Institute for Chemical Biology, where he continues to work in molecular recognition and self-assembling systems.

Fraser Hof was born in Medicine Hat, Alberta, Canada. He obtained a B.Sc. in chemistry at the University of Alberta in 1998, and is currently pursuing a Ph.D. under the direction of Professor Julius Rebek, Jr. at the Scripps Research Institute in La Jolla, California. He has been awarded postgraduate fellowships by the Natural Sciences and Engineering Research Council of Canada and the Skaggs Institute for Chemical Biology. His research focuses on the rational design of novel emergent properties through supramolecular chemistry.

Stephen Craig received his undergraduate degree in chemistry at Duke University, Durham (1991) and obtained an M.Phil. degree from Cambridge (1992) and a Ph.D. from Stanford University (1997). After two years as a Research Chemist in DuPontCentralResearch,hewenttoScrippsResearchInstitutein1999wherehewasanNIHpostdoctoralfellowinthelab of Professor Julius Rebek. He joined the Duke chemistry department in 2000 as an Assistant Professor, where his research interests center around the physical organic chemistry of materials.

Colin Nuckolls joined the faculty of Columbia University as an Assistant Professor of organic chemistry in July 2000. His group is studying the properties of materials that form through self-assembly. Previously, he was a National Institutes of Health post-doctoral fellow in the laboratory of Professor Julius Rebek, Jr. at the Scripps Research Institute. His undergraduate degree was awarded from The University of Texas, Austin where he worked with Professor Marye Anne Fox. He obtained his doctoral degree from Columbia University where he studied under the tutelage of Professor Thomas Katz.

1490 Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

Molecular Capsules

REVIEWS

Molecules that form supramolecular capsules are defined by two fundamental emergent properties: self-assembly and the encapsulation of guest molecules. Self-assembly is based on capsule components bearing complementary functional groups capable of reversible, noncovalent interactions. The noncovalent forces that are useful in constructing capsules are primarily hydrogen bonds and metal±ligand interactions. Both enjoy facile reversibility and reliable directionality, but hydrogen bonds offer greater plasticity and faster equilibra- tion, while metal ± ligand bonds typically offer greater strength and more rigidity. High-symmetry designs are used to multiply these individually weak and reversible interactions into coherent structures with lifetimes that range from microseconds to hours. The subsequent encapsulation of guest molecules is dependent on the complementarity of the guest×s size, shape, and chemical surface with the cavity of the host. The filling of space within the host is of utmost importance: nature abhors a vacuum, and this would seem- ingly include even those vacuums that measure only 1025 li- ters ![33]

Although these capsules are constructed with the express purpose of isolating guest molecules from the bulk solvent, the role of solvent in the formation of the capsules cannot be ignored. The medium must not disrupt the interactions that hold the components of the capsule together. Capsules constructed through metal±ligand interactions are typically disrupted by strongly ligating solvents, while they may remain stable in water. In contrast, solvent competition for hydrogen bonds prevents capsules constructed using these forces from being stable in aqueous media. The space-filling properties of the solvent must also be considered when dealing with encapsulation complexes. While the encapsulation of the solvent itself is sometimes desirable, the use of a large solvent that is physically excluded from the cavity can be an important tactic when encapsulating other guest molecules.[34]

2. Structural Motifs for Encapsulation
2.1. Glycoluril-Derived Hydrogen-Bonded Capsules

Glycoluril has been used to spectacular effect in the construction of supramolecular systems.[35] A more detailed discussion of the guest-binding properties of self-assembling glycoluril-based capsules than that which follows can be found in a recent review.[36]

Reversibly formed molecular capsules began with the TMtennis ball∫, 1 (Figure 1 a). The monomer (2) consists of two glycoluril subunits appended to a central aromatic skeleton; the glycoluril units provide curvature and a self- complementary hydrogen-bonding motif. The tennis ball is held together by eight hydrogen bonds, and as a host structure has a tiny cavity capable of housing guests with a volume of about 50 ä3. Accordingly, the tennis ball includes methane, ethane, ethylene, and the noble gases, while larger guests such as propane, allene, and isobutylene are excluded.[37, 38] Varia- tion in the spacer leads to smaller[39] and larger capsules (TMsoftballs∫, 3, Figure 1 b).[40±42] The same general symmetry remains, but apart from binding larger guests, the TMsoftballs∫

Figure 1. Self-assembling glycoluril-based dimeric capsules: a) the TMtennis ball∫ and b) the TMsoftball∫. As in most hydrogen-bonded capsules, curved monomers and self-complementary hydrogen-bonding seams are necessary components of the capsule geometry. (Some substituents and hydrogen atoms have been omitted for clarity.)

(their internal volumes lie between 240 and 320 ä3) are also capable of simultaneously binding two copies of moderately sized guests such as benzene.[41] Glycolurils were also ap- pended to spacers of threefold symmetry to form small, rigid (6, Figure 2 a)[43] and large, flexible capsules(8 a and b, Fig- ure 2 b).[44]

Cyclic sulfamides share the self-complementary hydrogen- bonding patterns of glycolurils. However, given the oppor- tunity, sulfonamides and glycolurils prefer heteromeric hydro- gen bonds. That is, they attract each other rather than themselves. A monomer such as 9 that contains both func- tional groups is programmed for self-assembly: if the groups appear at the ends of a suitably curved structure, assembly proceeds in a head-to-tail manner, with the best hydrogen donors and acceptors in contact.[45] The result is a capsule (10) made up of four subunits surrounding a cavity with a volume of about 160 ä3 (Figure 3 a).[46] The entropic penalty of bringing together four monomers and one guest in a single, discrete complex is forfeited by the enthalpic gains provided by the formation of 16 hydrogen bonds and whatever host ± guest interactions are on offer. As with the dimers, the tetrameric capsules bind molecules on the basis of size, shape, and chemical functionality. The larger monomer 11 assembles into the tetrameric capsule 12 (Figure 3 b) that encapsulates

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

1491

REVIEWS

J. Rebek, Jr. et al.

Figure 2. C3-symmetric self-assembling glycoluril-based dimers: a) the TMjelly donut∫ de- scribes a flattened cavity. b) Capsule 8 b contains holes through which small guests may pass freely. (For clarity here and in the following Figures, only hydrogen bonds in the foreground are depicted and some substituents and hydrogen atoms have been omitted.)

2.2. Cyclophane-Based Hydrogen-Bonded Supramolecular Capsules

Calix[4]arene and resorcin[4]arene are much admired, even standard subunits of self-assem- bled capsules. Both molecules exhibit variable conformations that can, through appropriate derivatization, be fixed into a single bowl-shaped conformation. The concave face of a bowl represents one half of a closed-shell topology, and a variety of functional groups mediate the corresponding dimerization. We refer the reader to a recent review[48] for a more detailed discussion of calixarenes than that which follows.

Secondary ureas were installed on the upper, wider rim of a calix[4]arene. In the presence of an appropriate guest molecule the ureas from two such calixarene monomers interdigitate and organize a directional seam of 16 hydrogen bonds around the equator of a dimeric capsule (14, Figure 4).[49] A variety of aromatic, aliphatic, and cationic guests are held within the twisted, bipyramidal cavity of approximately 180 ä3. The distal urea nitrogen atoms can be easily adorned with a variety of functional groups that alter the self-organizing behavior of the calix[4]arene monomer.[50, 51] This modularity programs hetero- dimeric assemblies (13 a,b),[52] kinetic stability (13 c,d),[53] and chirality (13 e)[54, 55] into the monomers. Larger calix[6]arene capsules have also recently been reported.[56]

The bowl shape of calixarenes and resorcinar- enes led to other versions, inspired by the report by MacGillivray and Atwood of a spectacular hexameric capsule in the solid state.[57] The structure shows a chiral arrangement of six resorcinarene subunits (15 a) enclosing an

Figure 3. Self-assembling tetrameric capsules. Two seams, each comprised of eight hydrogen bonds, stitch together a total of four identical monomers in a head-to-tail manner.

guests as large as the natural product longifolene within its expanded cavity (about 270 ä3).[47]

Figure 4. Self-assembling calix[4]arene tetraureas. The formation of ho- modimers (for example, 13a¥13a14a) and heterodimers (for example, 13 a ¥ 13 b) is determined by the identity of the urea substituents on the basis of electronic or steric properties.

1492

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

Molecular Capsules

REVIEWS

enormous cavity of about 1375 ä3 (Figure 5). The hexameric capsule 16 a features a total of 60 hydrogen bonds, in which 8 ordered water molecules are recruited to integrate the architecture. A similar water-bridged spherical structure was indicated by the 1H NMR spectrum of 15 b in benzene, but no evidence of specific guest encapsulation by the cubic hexamer 16b was reported. A hydroxy derivative, also a hexameric assembly in the solid state, also did not give any clues as to what was trapped inside.[58]

Figure 5. Hexameric structure observed in the solid state for methyl- substituted resorcin[4]arene 15 a. Eight water molecules are included in the hydrogen-bonding seam. Soluble monomer 15 b also forms this structure in solution in the presence of appropriate guests.

We revisited the solution characteristics of the resorcinar- ene monomer 15 b (Figure 5) more than a decade after Aoyama et al. had described the formation of 1:1 complexes of 15b with dicarboxylic acids in CDCl3 solution and subsequently with ribose, terpenes, and even steroids.[59] We found kinetically stable complexes of 15b formed in wet CDCl3 when suitable guests were available.[60] Guests such as large tetraalkylammonium and tetraalkylphosphonium salts reveal complexes with a host :guest ratio of 6 :1. The depen- dence on the guest size correlates nicely with the expected cubic hexameric structure observed by MacGillivray and Atwood in the solid state. Additional evidence suggests that the charged guests are encapsulated as ion pairs. When tetrabutylammonium bromide acts as the guest, enough space remains to concomitantly encapsulate a secondary neutral guest such as 4-phenyltoluene, thus three different species occupy the cavity. In the solid state, water-bridged dimeric capsules of 15 with small alkyl ammonium guests were characterized by Murayama and Aoki[61] as well as Rissanen and co-workers.[62, 63]

Following the synthetic and structural work of Cram et al.[26] and Dalcanale and co-workers,[64, 65] we devised and

synthesized 17, in which a vase-shaped cavitand structure presents four imide functions around its rim (Figure 6). The molecule dimerizes through bifurcated hydrogen bonds to form a capsule (18) about the size of a can of tennis balls (the molecular sort, see Figure 1 a).[66] The nonspherical shape of the cavity accommodates elongated guests and also promotes

Figure 6. Self-assembly of an imide-substituted cavitand into a dimeric capsule capable of binding elongated guests.

the pairwise selection of two simple aromatic compounds in an edge-to-edge manner. The selection depends strongly on the shape and size of each guest. Two molecules of benzene or two molecules of toluene are encapsulated simultaneously, while two molecules of p-xylene are not. In the presence of benzene, toluene, and p-xylene the capsule shows a strong (about 20:1) bias for the simultaneous binding of a benzene:p- xylene guest pair over the constitutionally isomeric toluene: toluene pair. Although the origin of this preference is not known, it demonstrates the stunning selectivity of molecular capsules.

Other resorcinarene platforms have also been developed. Chapman and Sherman explored the use of ionic hydrogen bonds in the generation of self-assembled capsules using a partially deprotonated hydroxy-substituted resorcinarene (19, Figure 7 a).[67] Kobayashi et al. have constructed a similar methylene-bridged resorcinarene scaffold functionalized with four carboxylic acids (Figure 7 b).[68] Here, 2-aminopyrimidine is used as a wedge-shaped hydrogen-bonding bridge that forms two hydrogen bonds with each of two carboxylic acids on neighboring molecules (20). Resorcinarene 21, which employs hydroxy and ester functional groups as hydrogen- bonding donors and acceptors, respectively, allows the assembly of supramolecular capsule 22 (Figure 7 c).[69] Flexi- ble electron-rich walls are able to collapse and form good contacts with a p-acceptor guest, which results in a dimeric capsule that binds a tropylium cation within its cavity. The glycoluril module[44] and resorcinarene module have been hybridized in the supramolecular capsule 24 (Figure 8).[70] The large interior volume (about 950 ä3) allows for the TMhost within a host∫ supramolecular encapsulation of ionic cryptate complexes, an arrangement analogous to Russian matryoshka dolls.

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

1493

REVIEWS

J. Rebek, Jr. et al.

Figure 7. Dimeric molecular capsules based on resorcin[4]arene building blocks: a) a dimeric capsule stitched together by charged hydrogen bonds, b)a capsule utilizing 2-aminopyridine as a hydrogen-bonding wedge (reprinted with permission from ref. [69]), c) the flexible walls of this resorcinarene collapse to give p contacts with encapsulated guest.

ideal hydrogen bonds requires molecules with curvature elsewhere, in the metal ± ligand assemblies metal hinges can be installed in the corners while the walls are constructed from flat ligand panels. Dalcanale and co-workers have created a variety of bridged resorcinarenes functionalized with four nitrile groups for the purposes of metal±ligand- directed self-assembly.[72±74] Square-planar palladium or plat- inum complexes having two labile ligands in adjacent positions can act as coordinating corner units for the self- assembly of such nitrogenous ligands. The resulting metal ± ligand interactions (and the assemblies that they generate) are reversible and robust in a variety of solvents. The combination of a C4v-symmetric resorcinarene ± nitrile ligand and a right- angle metal subunit in a 1:2 ratio leads to the formation of a self-assembled supramolecular cage complex 25 (Figure 9 a). The complex bears a total 8 charge, and the cavity is found to encapsulate one of the eight counterions. The assembled structure is stable in water, but is subject to decomposition by competing ligands such as triethylamine or acetate.

Shinkai and co-workers have demonstrated that the sub- stitution of pyridines into calix[4]arenes also results in the self-assembly of supramolecular capsules. A rigidified calix- arene monomer displaying four pyridine ligands undergoes metal-directed self-assembly to produce capsule 26 (Fig- ure 9 b) in a manner analogous to that of 25.[75] The bridging glycol substituents at the lower rim stabilize the C4v-symmet- ric cone conformation necessary for assembly. Modification of a homooxacalix[3]arene derivative to include 4-pyridyl groups gives a species that undergoes metal-directed self-

Figure 8. Self-assembly of a resorcin[4]arene ± glycoluril hybrid capable of encapsulating ion ± cryptate complexes.

2.3. Cyclophane-DerivedMetal±Ligand-Based Supramolecular Capsules

Hydrogen bonds do not have exclusive rights to the assembly of supramolecular capsules, and molecular housings built by using metal ± ligand interactions have their own architectural style.[71] Whereas the rectilinear arrangement of

The water-soluble p-sulfonatocalix[4]arene 30 is even more impressive in its response to small ligands and metals. It is coaxed to assemble by the presence of 1equivalent of pyridine N-oxide and 0.5 equivalents of La(NO3)3 .[81] The resulting spherical cluster (31, Figure11) has a delicate balance of hydrogen bonds, van der Waals forces, metal ± li- gand interactions, and electrostatic contacts that work in

1494

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

assembly to

form a supramolecular capsule (Figure 9 c).[76] This assembly (27) even encapsulates [60]full- erene (Ka 54 m1).

The list of metal ions useful for directing encapsulation continues to grow. Harrison and co-workers introduced tridentate chelating ligands as structural elements. A resorcinarene functionalized with four iminodiacetate groups (28) shows an affinity for binding CoII, CuII, and FeII salts. The result is the complex- ation of each metal in a chelated pseudo- octahedral environment and the generation of supramolecular capsules 29 (Figure10).[77±80] They are stable in water and encapsulate a wide variety of organic compounds, such as cyclic and acyclic aliphatic alcohols, ethers, ketones, esters, and halides, within a cavity of approximately 215 ä3. Compounds held in close contact with the metals that line the cavity experience enormous paramagnetic shifts. Guests encapsulated within the CoII complex show upfield chemical shifts of 18 to 40 ppm upon complexation.[80]

Molecular Capsules

REVIEWS

Figure 9. Metal-mediated dimerization of cyclophane-based ligands to form supramolecular capsules.

Figure 10. Dimerization of a cyclophane bearing four tridentate ligands mediated by four metal ions in pseudo-octahedral environments.

concert to create an ordered supramolecular capsule. X-ray crystallographic analysis shows: 1) 12 calixarene units func- tioning as pyramidal wedges, 2) a pyridine N-oxide molecule filling the cavity of each calixarene, and 3) the lanthanide ions

Figure 11. Colossal supramolecular capsule arising from self-assembly of 12 calix[4]arene subunits, 12 copies of pyridine N-oxide, and 6 La3 ions. One calix[4]arene subunit has been omitted to allow visualization of the capsule interior, which is occupied by 2 Na ions and 30 ordered water molecules. (Reprinted with permission from ref. [81].)

acting as coordinating hinges between calixarenes of adjacent clusters. The internal cavity of this spherical assembly has a prodigious volume of about 1700 ä3, and is occupied by an

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

1495

REVIEWS

J. Rebek, Jr. et al.

ordered cluster of 2 sodium ions and 30 water molecules. On the scale of the other capsules, this resembles a soccer ball. Under a different stoichiometry, the same three subunits can also form open-ended helical tubes in the solid state.[81]

2.4. Complexes Based on Tris(pyridine) and Tris(pyrimidine) Ligands

In a dramatic departure from cyclophane-based capsules, Fujita et al. have used simple triangular heterocyclic ligands (Figure 12) in combination with cis-enforced square-planar Pd and Pt complexes for the construction of highly symmetric supramolecular capsules.[82] The positively charged metal centers impart water solubility on the complexes, and their relatively hydrophobic cavities bind a variety of organic guest molecules. A brief overview of the striking structures born of this motif is now given. For a more detailed discussion we refer the reader to a recent review.[82]

Figure 12. Triangular pyridine/pyrimidine ligands for the construction of supramolecular capsules, and the palladium subunit (33) commonly used to stitch them together.

The simplest of these systems is created when a tris(pyr- idylmethyl) ligand (32) is combined with Pd-based corner unit 33 in the presence of a suitable organic guest.[83] Two ligands of 32 bind a total of three metal centers (Figure 13 a) to produce a C3v-symmetric supramolecular capsule (38). The palladi- um ± pyridine bonds are stable in protic solvents and the high overall charge (6) of the complex imparts water solubility. The hydrophobic interior of the capsule is aptly filled by organic anions such as adamantanecarboxylate. The subunits aggregate into an uncharacterized oligomeric state in the absence of a suitable guest.

Analogous rigid planar threefold-symmetric ligands form higher order geometric structures. Ligand 35, with three 4-pyridyl subunits around a central triazine core, forms supramolecular capsule 41 in the presence of a cis-protected square-planar Pd or Pt subunit (Figure 14 a).[84] The metal atoms reside at each corner of an octahedron with the longest

Figure 13. a) A supramolecular capsule built from two flexible ligands and three metal subunits. b, c) Self-assembly of a C3-symmetric ligand into a bowl-shaped structure which undergoes a hydrophobic dimerization. Four copies of m-terphenyl are encapsulated (shown as CPK models). (Re- printed with permission from ref. [89].)

closed by the capsule about 500 ä3. The platinum-based capsule 41c is remarkably stable[85] and encapsulates several guests the size of adamantane.[86] The encapsulation of four copies of each guest takes place in a cooperative manner that is independent of the nature of the guest. In contrast, tris(pyrimidine) ligand 36 forms a hexahedral supramolecular capsule when combined with a small excess of PdII complex 33 (Figure 12).[87] The self-assembly of this hexahedral structure entails recognition and binding among a total of 6 triangular ligands and 18 metal ions. Unlike the previous structure (41) in which planar ligands filled alternating faces of a polygon, each face of the hexahedral capsule 42 is completely enclosed by the planar threefold-symmetric ligands (Figure 14 b). The volume of the cavity enclosed by this capsule is considerable (about 900 ä3), but only water or small gas molecules may pass through the meager pores (2 2 ä) in the structure. Ligand 37, a variant of 36, also gives rise to a capsule with hexahedral geometry analogous to that of 42 (Figures 12 and 14), but with some vacant metal binding sites that result in formation of hydrophobic clefts.[88] These apertures allow for the encapsulation of small molecules such as CBr4 , a behavior that is not displayed by the parent capsule.

The 3-pyridyl-substituted ligand 34 is unable to form a closed-shell topology with organometallic corner subunits. Instead, four copies of ligand 34 and six copies of Pd-subunit 33 self-assemble into an open bowl-shaped structure 39 (Figure 13 b).[89] These hemispherical superstructures assem- ble in water into discrete dimeric supramolecular capsules (40) in the presence of large aromatic guests (Figure 13 c). X-ray crystallographic studies reveal that four copies of m- terphenyl or six copies of cis-stilbene form compact hydro-

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

metal ± metal separation being 1.9 nm and 1496

the

volume

en-

Molecular Capsules

REVIEWS

Figure 14. a) A supramolecular capsule constructed from 4 triangular ligands and 6 metal units. The metal atoms define the corners of an octahedron and the ligands occupy alternating faces. b) A hexahedral supramolecular capsule composed of 18 metal ions and 6 triangular ligands. (Reprinted with permission from ref. [87].)

phobic clusters that are encapsulated by the discrete dimeric superstructure. Only dispersive forces and the hydrophobic effect act to hold the capsule together–there is no direct metal ± ligand bonding between the supramolecular bowls that comprise the capsule halves. In the absence of direct contacts between the molecules that make up the bowls, it is the guests–bound within both bowls–that provide the bridging interactions that drive dimerization.

The metal-induced self-assembly of non-C3-symmetric ligand 43 (Figure 15 a) is guest-dependent.[90] An open cone structure made up of four ligands and eight metals is formed in the presence of benzil. More remarkable is that the addition of CBr4 drives the assembly of the four triangular ligands and eight metal centers into a closed-shell tetrahedral capsule 44 (Figure 15 a). In this case the ligands are arranged in a head- to-tail manner, and fill each face of the tetrahedron so as to completely surround the encapsulated guest.

Figure 15. Self-assembly of tetrahedral supramolecular metal ± ligand clusters: a) four triangular ligands 43 occupy the faces of the tetrahedron containing eight metal ions (reprinted with permission from ref.[88]), b) four metal ions define the corners of the tetrahedron and six ligands span the edges, c) two other bis(catechol) ligands that form tetrahedral capsules in a manner analogous to ligand 45.

2.5. Tetrahedral Metal ± Ligand Clusters as Supramolecular Capsules

Other motifs have been explored for the creation of tetrahedral ligand ± metal clusters. Saalfrank et al. have re- ported a family of M4L6 capsules of tetrahedral symmetry,[91, 92] and have also shown one example that encapsulates its NH4 counterion.[93] Other research groups have achieved similar results with different metal ± ligand combinations, the encap- sulated species being either a counterion or adventitious solvent molecule.[94±99]

The rich host ± guest chemistry of tetrahedral metal ± ligand clusters has been beautifully developed by the research group of Raymond. In their system, six bis(catechol) ligands span the edges between four metal atoms that define the corners of a tetrahedron. Most bridged bis(catechols) form M2L3 heli- cates as the thermodynamically most-favored framework, but here the structure of the spacer between the catechol ligands instructs the assembly. Bis(catechol) 45 is linked with a spacer that encourages the adoption of a rigid C2-symmetric con- formation. The beautiful self-assembly of M4L6 tetrahedral

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

1497

REVIEWS

J. Rebek, Jr. et al.

clusters 46 is observed when the ligand is combined with TiIV, GaIII, or FeIII ions (Figure 15 b).[100, 101] The capsule bears a net 8 charge, and reveals an aptitude for binding positively charged guests. The tetrahedral clusters select tetraalkylam- monium guests on the basis of size, yet are also capable of adjusting their cavity volumes from about 200 to 300 ä3 in response to guests of different sizes.[102] Like 24, they form complexes-within-complexes through the encapsulation of small alkali ions held by crown ethers.[103] The pyrene-bridged ligand 47 (Figure 15 c) self-assembles into a tetrahedron only in the presence of guests, which are bound in a manner analogous to that with 45,[104] while the anthracene-bridged ligand 48 displays more intricate behavior.[105] In the absence of guests, the combination of 48 with appropriate TiIV or GaIII subunits results in the formation of the M2L3 helicate as the most stable supramolecular structure. The addition of tetramethylammonium ions to the mixture, however, orches- trates a spontaneous rearrangement to the capacious M4L6 tetrahedral cluster. Here, the binding of a relatively small guest molecule provides the thermodynamic bias for a wholesale reorganization of six large organic ligands and four metal centers. More complex metal ± ligand systems based on multiple copies of two different metals bridged by ditopic ligands have also been developed, but their guest encapsula- tion properties have not yet been fully explored.[106]

3. The Ins and Outs of Encapsulation Complexes 3.1. Guest Dynamics and Behavior

Encapsulation places obvious constraints on the transla- tional motion of the guest molecule. Bˆhmer and co-workers provided some of the earliest understanding of dynamics within the calix[4]arene ± tetraurea capsules 14 (Figure 4). The diffusion coefficients of encapsulated and free guest molecules were determined by the pulse gradient spin echo (PGSE) NMR technique.[107] The diffusion coefficients of the encapsulated guests were found to match well with those of the assembled host, while those of the free guest were much higher.

The effects of encapsulation on the guest×s rotational freedom is usually apparent from the NMR spectra. Benzene tumbles rapidly and p-xylene slowly on the NMR timescale in the cylindrical capsule 18 (Figure 6); toluene shows a broad- ened signal characteristic of an intermediate tumbling rate. In the softball 3 b (Figure 1) with encapsulated [2.2]paracyclo- phane, 13C spin-lattice relaxation studies gave evidence of correlated bulk host ± guest movement. This large guest is apparently wedged into a limited space.[108]

The research group of Bˆhmer took advantage of hydro- gen-bonded calixarene capsules (14) with decreased symme- try to study the rate of capsule dissociation (0.26 s1) and the rate of guest exchange (0.47 s1) using 1H NMR NOESY experiments.[109, 110] The intimate relationship between the host and guest is also reflected in the determination of the geometry and dynamics of an encapsulated guest through NMR spectroscopic analysis of the host alone. The overall symmetry of the host complex is affected by the binding

orientation and dynamics of the guest molecule bound within the cavity. In one elegant example by Fujita and co-workers, simple 1D NMR studies on host 41 (Figure14) yielded detailed information on both the binding orientation and the temperature-dependent dynamics of the included guest.[111]

Effects on the guest×s internal molecular dynamics are also directly measurable. The ring inversion of cyclohexane within capsule 6 (Figure 2) was studied through the use of [D]11- cyclohexane as a guest molecule.[112] The barrier to ring inversion within the flattened cavity of 6 (10.55 kcal mol1) is increased by 0.3 kcal mol1 relative to the value found in free solution (10.25 kcal mol1) ; ground-state stabilization through CH-p interactions within the cavity is thought to be the cause for this difference. The encapsulation of [D]11cyclohexane within a calixarene ± tetraurea capsule (14, Figure 4) that described a roughly spherical cavity resulted in no observable change in the barrier to ring inversion. Conversely, the internal dynamics of 1,4-dioxane and 1,4-thioxane encapsu- lated within capsule 19 (Figure 6) are significantly restrict- ed.[113] Despite the pseudo-spherical nature of the cavity surrounding the guests, the barrier to conformer interconver- sion is increased by a relatively large 1.6 ± 1.8 kcal mol1 upon binding. In general, these experiments suggest that the internal dynamics of included guests are controlled specifi- cally by host ± guest interactions.

3.2. Control of Guest Release

For the development of encapsulation-based applications it is more urgent to control guest exchange than to understand it. The guests within most hydrogen-bonded capsules are liberated by the addition of solvents that compete effectively for hydrogen bonds. The dissociation is thermodynamically and kinetically facilitated by these competitive solvents. Capsules based on metal ± ligand interactions are in general subject to decomposition (and concomitant guest release) by the addition of strong nucleophiles and/or subjection to elevated temperatures. Although these environmental changes do successfully bring about the liberation of guests, there is a need for more specific and reversible methods for the control of guest encapsulation and release.

A convenient stimulus for the reversible control of many metal ± ligand based encapsulators involves changes in pH values. Harrison and co-workers have used pH changes to reversibly trigger the self-assembly of resorcinarene ± imi- nodiacetate capsules and accompanying guest encapsula- tion.[77] The cobalt-based capsule 29 (Figure 10) is assembled at pH 6, but exists in a monomeric state at pH 1. The exchange between these two states is reversible. The uptake of gases in the tennis ball 2 c (Figure 1) has also been controlled by the action of acids and bases on the peripheral amino groups.[39]

Instead of altering the structural components of the capsule, guest release can be accomplished most simply by guest exchange. If the system is under thermodynamic control (as are most self-assembled capsules) then the weakly held guest is displaced by the strongly held one. One example of this type of supramolecular substitution reaction is the exchange of adamantane with paracyclophane in the softball 3b (Fig-

1498

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

Molecular Capsules

REVIEWS

ure 16).[114] At the millimolar concentrations convenient for study by NMR spectroscopy the exchange takes place with a half-life of about one hour and the process has much in common with conventional substitution reactions. At low concentrations of the incoming guest the slow step is SN2-like, while at high concentrations of the incoming guest the slow step is SN1-like. In both cases, complete dissociation of the capsule is not necessary, as this would rupture all 16 hydrogen bonds. Instead, a lower energy process is proposed that involves opening TMflaps∫ on the softball×s surface (Figure 16). This exposes the resident guest to the incoming TMnucleophile,∫ which is either the solvent or new guest. As a consequence the guest exchange is faster than dissociation (see Section 4.1).

Figure 16. A two-step substitution reaction in a supramolecular capsule. The solvent-filled capsule is intermediate in the replacement of adaman- tane by the thermodynamically more favored [2.2]paracyclophane. Parti- ally open capsules are proposed as transition states (S solvent).

4. Form to Function

Supramolecular chemistry has matured to a degree that the design and synthesis of molecules that self-assemble into predictable supramolecular structures is becoming routine. Much of this research is curiosity driven, but the application of self-assembling systems to the development of functional devices should not be ignored, particularly in this, the TMnano∫ decade. The development of functional properties from self- assembly is merely the first step towards this goal.

4.1. Chirality in Encapsulation Complexes

Chiral supramolecular complexes are always popular topics, especially when noncovalent interactions direct the assembly of achiral components into chiral superstructures. In the absence of a chiral bias the structures appear as racemates, but with chiral information present in the system, the spontaneous formation of an excess of the appropriate supramolecular structure can be the outcome.[115, 116] Several studies have taken advantage of the intimate relationship between encapsulated molecules and the supramolecular

capsules that bridle them, thus granting a unique perspective on the transfer of information within complexes that is governed by noncovalent interactions. At the outset, the larger distances (compared to covalent bonds) and the flexibility of the weak, often nondirectional forces, did not guarantee success.

The tendency for calixarene ± tetraurea monomer 13 b functionalized with aryl sulfonamides to exclusively (b 98 %) form heteromeric capsules with monomer 13 a, functionalized with a simple aryl group, has previously been discussed in Section 2.2.[52] When heteromeric dimers such as 13 a ¥ 13 b form, the cyclic directionality of the urea hydrogen bonding seam results in the generation of racemic chiral species (Figure 17 b).[54] The head-to-tail arrangement of the

Figure 17. Chirality in calix[4]arene ± tetraurea capsules : a) cyclic direc- tionality of the urea hydrogen-bonding seam, b) capsules constructed from two different achiral components (for example, 13a¥13b) are formed as equilibrating cycloenantiomers, c) the presence of chiral groups on one of the two subunits can induce complete selectivity in urea directionality to produce an optically pure capsule (for example, 13 a ¥ 13 e).

urea units at the equator can be clockwise or counterclock- wise, given a reference point of the poles. Interconversion of these enantiomers occurs through the rotation of functional groups that make up the hydrogen-bonding seam, or through complete dissociation and recombination of monomers. Either mechanism would require the eventual breaking of all 16 hydrogen bonds, and as such the reversal of hydrogen- bonding directionality (and thus interconversion of enan- tiomers) is slow on the NMR timescale.[109] Chiral guests, however, were not capable of significant differentiation of the two resulting enantiomeric capsules 13 a ¥ 13 b.[54]

A bias was introduced with chiral auxiliaries attached to the distal urea nitrogen atom. A screening of amino acid derivatives led to the observation that calixarenes appended in this manner with ˜-branched amino acids have a predi-

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

1499

REVIEWS

J. Rebek, Jr. et al.

lection for association with calixarenes bearing aryl-substi- tuted urea groups.[55] Mixing the valine-derived monomer 13 e and aryl monomer 13a favors the assembly of the hetero- dimeric capsule 13 a ¥ 13 e almost exclusively (b 98 %). This capsule is analogous to the capsule 13 a ¥ 13 b formed from achiral components. Here, the presence of chiral groups on one subunit results in complete asymmetric induction of the capsule×s cycloenantiomerism (Figure 17 c). The resulting enantio- and diastereopure capsule 13 a ¥ 13 e shows an approximately 13% excess of one diastereomeric complex for the binding of the chiral guest norcamphor from a racemic mixture. The chiral auxiliary groups are not in direct contact with the encapsulated guest molecule. Instead, their influence is transmitted to the guest through the directionality imparted to the hydrogen-bonding seam that lines the cavity.

Glycoluril-based monomers (Figure 1) have also been employed in the construction of chiral self-assembled cap- sules. The simple monomers (1, 3) contain two mirror planes, which are both preserved in the dimeric assembled state. Analogous monomers lacking one of the two mirror elements have been synthesized.[117] These monomers are achiral, but self-assemble into dimeric supramolecular capsules that retain no mirror planes in the assembled state (Figure 18 a).

Figure 18. TMSoftball∫ monomers possessing only one mirror element spawn dimeric structures lacking any mirror symmetry: a)the host is templated with a chiral excess by the binding of a chiral guest, b) host racemization is slow following removal of the chiral guest and the enantiomeric excess of the capsule persists for several hours.

Again, the chiral capsules formed from these achiral compo- nents are formed as an equilibrating racemic mixture. The use of monomer 49, which incorporates symmetry-breaking elements adjacent to the encapsulated guest, results in a host ± guest pair capable of significant transfer of chiral information.[118, 119] The binding of an enantiomerically pure guest can bias the self-assembly process (a form of imprinting) such that one enantiomeric capsule is favored over the other by as much as a factor of four. In a reversal of the flow of chiral information, this diastereomeric host ± guest complex can then be used for noncovalent chiral templating (Fig- ure 18).[120] In this procedure, an optically pure guest is used to imprint the formation of a single chiral softball enantiomer and is then rinsed out rapidly by an excess of an achiral guest or solvent molecule. Since the exchange of guests in glycoluril- based capsules is much faster than the dissociation of the capsule, the exchange occurs without racemization of the capsule itself. The ghost of the chiral guest allows the chiral capsule to discriminate between guest enantiomers for several hours before it returns to its thermodynamically determined state.

A recent study by Shinkai and co-workers has revealed a capsule that possesses helical chirality.[121] Monomer 50 (Fig- ure 19) undergoes metal-directed self-assembly to form a dimeric capsule analogous in structure to 27 (Figure 9). The decreased angle of the ligating atoms in 50 (120V) relative to 27 (180V) results in a trans-substituted metal subunit being utilized for the formation of a closed-shell dimeric assembly (51). The binding of alkali metal ions at the oxygen-rich lower rim of homooxocalix[3]arenes is known to affect the geometry and binding properties of the capsules constructed from such building blocks.[76] In the case of 51, the binding of Na ions

Figure 19. A homooxocalix[3]arene monomer forms helically chiral di- meric capsules in the presence of suitable metal subunits.

1500

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

Molecular Capsules

REVIEWS

induces a dramatic conformational shift that yields a helical arrangement of the capsule walls down the long axis of the capsule (Figure 19). Both M and P enantiomers are formed, as revealed by experiments with chiral shift reagents. As in the softball, chiral guests are able to induce a bias (up to 55 % for the encapsulated (S)-2-methylbutylammonium ion) in favor of one helical enantiomer over another.

In another example, the direction of chiral information flow is reversed. Derivatization of compound 9 (Figure 3) with a single hydroxy group provides a new monomer (52) that is chiral and resolvable into single enantiomers (Figure 20).[122] Self-assembly occurs spontaneously in the presence of an appropriate guest, and the enantiopure capsule (53) exhibits enantioselective (d.e. 60%) binding of chiral guests from a racemic mixture. The interaction of a single copy of monomer 52 with a single chiral guest molecule is presumed to offer a small energetic differentiation, but significant chiral recog- nition emerges when multiple copies of monomer 52 form a closed chiral space.

Figure 20. The self-assembly of an enantiopure monomer yields a chiral tetrameric capsule capable of discriminating between guests on the basis of chirality.

A common origin of chirality in metal-based supramolec- ular complexes is the h/v helical chirality associated with the octahedral arrangement of three bidentate ligands around a metal center. The tetrahedral clusters of Raymond and co- workers[101] demonstrate the spontaneous generation of supra- molecular chirality within self-assembled capsules based on metal ligation. The chirality of the four metal centers present in each tetrahedral cluster is strongly coupled by the bis(catechol) ligand 45 (Figure 15); the hhhh and vvvv clusters are formed to the exclusion of clusters of mixed configurations. The two resulting diastereomeric complexes formed when the ()-N-methylnicotinium cation is encapsu- lated by GaIII complex 46 (Figure 21) can be easily separat- ed.[123] Removal of the chiral guest by replacement with the

Figure 21. a) Facile racemization between the h and v configuration of chiral octahedral gallium(iii) catecholates. b)An optically pure self- assembled tetrahedron containing four such metal centers in a mechan- ically linked framework is isolated through encapsulation of a chiral guest, and is stable to racemization even after extended heating.

achiral Et4N ion yields the enantiopure tetrahedral clusters 46¥Et4N. The mechanical coupling of metal centers within the tetrahedral framework not only favors the presence of four homoconfigurational metal centers in a cluster, but also confers to each metal center a remarkable resistance to racemization. The racemization rate for isolated tris(catecho- late) ± GaIII centers is fast (10 s1) at room temperature. In contrast, an aqueous solution of (hhhh)-46 ¥ Et4N remains enantiopure after eight months at room temperature, and even extended heating does not induce racemization.

4.2. Dynamic Libraries of Molecular Receptors

Great diversity can be generated through the combination of relatively few components and a large library of multicom- ponent species can be developed. Lehn has pioneered the field of dynamic combinatorial chemistry, in which the com- position of equilibrating libraries of molecular receptors is trained by the presence of the desired target.[124±127] Molecular capsules are suitable candidates for this treatment. As such, the selection of molecular receptors from a library under thermodynamic control has been achieved by both hydrogen- bonded and metal ± ligand based supramolecular capsules.

Our efforts involved the monomer 9 that self-assembles to form hydrogen-bonded tetrameric capsule 10 (Figure 3). The variation of the substitution patterns at the central aromatic carbon atom gave a total of five complementary (and self- complementary) subunits (9, 52, 54 ± 56, Figure 22). The added functionality doesn×t disrupt the forces responsible for the assembly but does modify the size, shape, and chemical lining of the cavity. The five monomers represent a library of 613 possible tetrameric capsules with 70 different composi- tions.[128] The mixture was monitored by using electrospray

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

1501

REVIEWS

J. Rebek, Jr. et al.

Figure 22. A dynamic library of self-assembled tetramers arises from the combination of five complementary monomers. Each letter indicates a mass spectral peak corresponding to a unique molecular host that forms in varying quantities in response to the presence of different guest molecules. The composition of the library is biased towards the hosts that interact most favorably with the added guest.

mass spectrometry, a method that has recently come of age in the characterization of supramolecular complexes.[129, 130] The distribution of tetrameric capsules was measured by taking advantage of mass-labeled monomers. The composition of the mixture depends strongly on the nature of the guest: the receptors that best fit the structure of the added guest molecule spontaneously emerge as the predominant species in solution.

In another example, a variant on the tris(pyridine) trian- gular ligands (Figure 12) developed by Fujita and co-workers, is the set of supramolecular capsules arising from mixing one 4-pyridyl and two 4-pyridylmethyl subunits as ligands with a single metal unit as an adhesive. The C3 symmetry of the ligand (57) is broken, and metal-mediated self-assembly gives two isomeric capsules (Figure 23). The formation of the two receptors is strongly guest-dependent: one is favored by aromatic guests, while the other is favored by more spherical guests such as CBr4.[131] The diversity of receptors present in the library was increased by the addition of another ligand, 32. Equilibration within the resulting library of four receptors was efficiently controlled by the presence of suitable guest species (Figure 23).[132]

4.3. Capsules as Sensors

One function of supramolecular capsules is in small- molecule sensing, and we use an example of calixarenes to show how a binding event gives a detectable signal: a common transduction path is used for the detection of analytes at low concentration. Calixarene±tetraureas 58 and 59 were sub- stituted[52] at the lower rim with fluorescent dyes (Fig- ure 24).[133] The self-assembly and encapsulation processes of these monomers proceed in the usual manner with hetero- meric capsule 60 preferred to the formation of each homo- dimer (as in 13 a ¥ 13 b). The dyes were selected such that the emission spectrum of one dye (the donor, 58) overlaps the excitation spectrum of the other (the acceptor, 59). When the two dyes are in proximity to one another, excitation of the donor results in significant fluorescence resonance energy transfer (FRET) to the acceptor, and emission at the accept- or×s emission wavelengths is observed. The result is a FRET signal that is dependent on the assembly of 58 and 59. Nonspecific aggregation is negligible at the nanomolar con- centrations of the experiment, and no FRET is observed in the absence of a suitable guest species. In the presence of a

Figure 23. A dynamic library of four receptors arising from the combination of two ligands in the presence of a suitable metal subunit. Spherical guests such as CBr4 or CBrCl3 give rise to different hosts than include aromatic guests such as benzene or p-xylene. (Reprinted with permission from ref. [132].)

1502 Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

Molecular Capsules

REVIEWS

Figure 24. Encapsulation-dependent sensing of a guest. The presence of a guest encourages capsule dimerization, which in turn produces an optical signal through fluorescence resonance energy transfer between dyes covalently attached to each monomer.

suitable guest, encapsulation occurs and a FRET signal arises. In this way the encapsulation-dependent detection of 3-meth- ylcyclopentanone was achieved by the combination of 58 and 59 in p-xylene.

4.4. Self-Assembled Capsular Polymers

New structural and physical properties emerge from the polymeric assembly of compound 61, produced by covalently coupling a calixarene ± tetraurea monomer at the lower rim to another calixarene ± tetraurea monomer (Figure 25 a).[52, 134] The use of a rigid linker creates divergent tetraurea recog- nition elements that are unable to bond in an intramolecular sense. Instead, the encapsulation-driven self-assembly results in a polymeric chain of capsules (Figure25b). Like other polymers, TMpolycaps∫ display new properties on the macro- scopic scale. Unlike traditional polymers, the chains use reversible interactions and are formed under equilibrium conditions. Polarized light microscopy studies on concentrat- ed solutions of 61 reveal that the polymer displays a nematic liquid crystalline state, that is, the polymer chains self- organize in a linear array.[135] Chiral nematic (cholesteric) liquid crystalline phases emerge from analogous chiral monomers. Fibers that are pulled from the liquid crystalline melt also display order under a polarized light microscope, a behavior that is, perhaps, responsible for their surprising strength (crude measurements show that these fibers have a yield stress within an order of magnitude of that of covalent polymer fibers such as nylon-6). These samples also display bulk viscoelastic properties related to their polymeric nature.[136]

Figure 25. Encapsulation-dependent polymerization and gel formation. Monomer 61 forms polymeric chains of capsules (polycaps) in the presence of a suitable guest. Addition of noncovalent cross-linker 61 results in the formation of physical gels.

The doping of monomer 61 with a compatible cross-linking species (62) has dramatic effects on the bulk properties of the mixture.[136] Compound 62 relies on encapsulation to form self-assembled cross-links between polymer chains. Cross- linker 62, when present in concentrations as low as 5% relative to 61, causes the formation of a gel phase. These gels are reminiscent of conventional physical gels, in which covalent linear polymers are cross-linked by weak noncova- lent interactions. Here, the structural components are re- versed, but the viscoelastic behavior of the gels is quite similar. For example, the gels are dilatant: their viscosity increases with the application of shear. This result points to increased ordering under anisotropic flow, and ordered structures have been observed in surface transmission elec- tron microscopy images obtained from frozen chloroform.

In every case, the bulk properties of the polycaps are dependent on the presence of a suitable guest species and a solvent that does not compete for hydrogen bonds. The formation of polymeric chains, liquid crystalline phases, viscous polymeric solutions, and gels are properties funda- mentally derived from, and dependent on, molecular encap- sulation.

5. Control of Reactivity through Encapsulation

The most direct illustration of the effects of encapsulation on small molecules is the mediation of chemical reactions. Encapsulated molecules are removed from a sphere of solvation and placed in enforced proximity to the host and, if space permits, to other guests. This constrained environ- ment governs the guest×s encounters with potential reactive partners, as well as fundamentally altering the concentration (molecules per volume) of reactive species. This influence is exerted for the duration of the lifetime of the encapsulation complex, which may be from microseconds to hours.

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

1503

REVIEWS

J. Rebek, Jr. et al.

5.1. Acceleration and Catalysis

Two different approaches to supramolecular catalysis through encapsulation have been reported in the literature: 1) bimolecular catalysis can occur when two reactive partners are bound within a single capsule and 2)phase-transfer catalysis can occur when the capsule transports guests from one solvent phase to another. Both rely on turnover, the release of product (or passenger) and the re-uptake of reactants. Reversibility is the key to this behavior.

Initial reports of the self-assembled softball (3, Figure 1)[40] and its propensity for the simultaneous encapsulation of two guest molecules[41] raised the possibility of catalyzing a bimolecular reaction by encapsulation. Rate acceleration through encapsulation was observed in the Diels ± Alder reaction of benzoquinone (63) and cyclohexadiene (64) mediated by 3 b (Figure 26 a).[137] In the resting state of the systemtwomoleculesofbenzoquinoneareencapsulated

Figure 26. Diels ± Alder reactions mediated by a self-assembled capsule : a) the reaction of benzoquinone and cyclohexadiene is accelerated through encapsulation, but product inhibition prevents catalytic turnover, b) the reaction of benzquinone and 66 is accelerated by the capsule, and subsequent dissociation of product results in catalytic turnover.

strongly by 3 b : neither the capsule containing only cyclo- hexadiene nor the mixed encapsulation complex can be observed by NMR spectroscopy. Nevertheless, an encapsula- tion-dependentrateaccelerationofnearly200-foldoccursin the Diels ± Alder reaction between the two substrates. The rate acceleration likely arises from a mixed encapsulation complex, the counterpart of the Michaelis complex.[138] Derivatives of 63 and 64 that are not of appropriate size for encapsulation show no rate acceleration. Addition of a nonreacting guest that competes effectively for the catalytic site (an inhibitor) also prevented rate acceleration. Unfortu- nately, the product (65) of the reaction of 63 and 64 is a good

guest for the capsule, and strong product inhibition prevents turnover, that is, catalysis (Figure26a). The reaction of 66 with 63 was examined in the expectation that the loss of SO2 from the Diels ± Alder adduct would result in product (67) release and catalyst turnover (Figure 26 b). Instead, it was found that loss of SO2 does not occur under the reaction conditions. Nevertheless, the product (67) is fortuitously ejected from the softball by the quinone and catalytic turnover is the outcome.[139]

A different approach to encapsulation-dependent catalysis uses capsules constructed through metal ± ligand interactions operating in aqueous environments. Capsule 41a is highly charged (12), yet shows a propensity for encapsulating a variety of neutral hydrophobic molecules. This combination of properties make 41a a unique candidate for the encapsu- lation-dependent phase-transfer catalysis of reactions in water. The effect of 41a on the Wacker oxidation of styrene catalyzedby[Pd(NO3)2(en)](enethylenediamine)was examined to test this hypothesis (Figure27).[140] In the absence of capsule 41a, the transformation of styrene to acetophenonecatalyzedby[Pd(NO3)2(en)]proceedsonlytoa small extent in water (4%) as a result of the substrate×s low solubility in aqueous media. Under the same conditions, the

Figure 27. Reverse phase-transfer catalysis of the Wacker oxidation of styrene by a supramolecular capsule.

presence of a catalytic amount of 41a results in a dramatic increase in the yield of acetophenone (82%). Experiments carried out in the presence of styrene and a competing guest (1,3,5-trimethoxybenzene) decreased product formation dra- matically (3%). The structural PdII components of capsule 41 a alone are not sufficient for catalysis of the reaction ; in all casestheadditionofcatalytic[Pd(NO3)2(en)]isnecessary. The result is a unique TMdouble catalyst∫ system wherein 41a acts as a phase-transfer catalyst and a separate PdII species acts as the oxidation catalyst (with reoxidation by air). This reverse phase-transfer methodology can be extended to other substituted styrenes that are encapsulated by 41a.

In another example, complex kinetic behavior reminiscent of autocatalysis is observed when reagents are compartmen- talized within a supramolecular capsule.[141] Dicyclohexylcar-

1504

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

Molecular Capsules

REVIEWS

bodiimide (DCC) is encapsulated by cylindrical capsule 18 (Figure 28). Addition of benzoic acid 68 and aniline 69 to the mixture gives rise to complex kinetic behavior. Trace amounts of DCC free in solution promotes the formation of an amide bond between the acid and amine reactants. The products of this reaction are anilide 70 and dicyclohexylurea (DCU), both of which are better guests for capsule 18 than DCC. Accordingly, increasing amounts of DCC are displaced from the capsule by 70 and DCU, and the rate of the reaction increases as the reaction proceeds. The kinetics possess a sigmoidal character that depends on product formation, yet this is not classical autocatalysis, as there is no single catalyst in the system. The nonlinear kinetics can be viewed as an emergent property of the system as a whole, with the partnership of compartmentalization and molecular recogni- tion giving rise to chemical amplification. This result, while not easily classified, highlights the role that compartmental- ization may play in the creation and maintenance of complex systems.

5.2. Stabilization of Reactive Species

Reversibly formed capsules have successfully stabilized species that are not otherwise stable in free solution. The processes that are responsible for self-assembly and encapsu- lation can provide enough free energy for an encapsulated guest to alter its own internal equilibria. In the simplest of these cases, encapsulated guests display altered conforma- tional preferences. Supramolecular capsules can also stabilize encounter complexes formed between multiple guests that are not otherwise observed. In the most dramatic examples, encapsulation can be used for the stabilization and isolation of reactive intermediates.[142]

As described in Section 2.2, self-assembled capsule 18 is capable of encapsulating long cylindrical molecules that are complementary in both shape and size to the dimensions of its elongated cavity.[66] A preference for the binding of trans- stilbene over cis-stilbene underscores this selectivity.[143] The binding of N-methylbenzanilide (71) demonstrates a more subtle set of characteristics imparted by the host capsule. Although 71 is known to prefer the E conformation in free

solution, the physical constraints provided by capsule 18 force 71 to adopt the unfavored Z conformation (Figure 29).[143] Like all properties governed by encapsulation, the prefer- ences imparted upon 71 arise from a combination of equilibrium processes. The self-assembly of the capsule, the encapsulation of the guest, and the conformational state of the guest molecule are all dynamic processes that conspire to yield the end result–in this case a simple conformational shift.

Figure 29. A shift in conformational equilibrium (E)-71b(Z)-71 brought about by shape-selective encapsulation within an elongated capsule.

Kusukawa and Fujita used capsules constructed with metal ± ligand interactions for the stabilization of unfavored conformations. The treatment of an aqueous solution of capsule 41a with a solution of 4,49-dimethylazobenzene (72, cis:trans 1:6) in hexane results in the formation of an unusual complex within the capsule walls (Figure 30 a). The capsule selectively binds two equivalents of cis-72.[144] 2DNMR studies using the analogous cis-stilbene 73 shows NOE contacts between the vinyl protons and methyl protons of the guest, which provides additional evidence for the pro- posed dimeric guest cluster. The cis-azobenzene molecules are considerably stabilized within this encapsulation complex: exposing the solution to visible light for several weeks did not result in the production of any of the thermodynamically favored trans-azobenzene. Molecular modeling studies reveal that the dimeric hydrophobic guest complex is too large to have formed outside the capsule and entered as a single species. Since the structural elements of the capsule are not equilibrating under the conditions of these experiments, the

Figure 28. Compartmentalization of reactants and products gives rise to emergent nonlinear behavior. DCC is initially sequestered from the reaction medium through encapsulation within cylindrical capsule 18. As the reaction between acid 68 and amine 69 proceeds, the coupled product 70 and DCU are generated and both displace DCC from the reversibly self-assembled capsule.

hydrophobic dimer must form within the capsule walls.

This TMship-in-a-bottle∫ encap- sulation process is not unique to azobenzene and stilbene guests. The exposure of Pt-based cage 41 c to phenyltrimethoxysilane results in the formation of a cyclic silanol trimer 74a within the cavity of the supramolecular capsule (Figure 30 b).[145] Cyclic trisilanol oligomers of this type are considered to be intermedi- ates in the sol±gel condensa- tion. Although the cyclic tetra- mer has been prepared by other methods, the highly reactive cy-

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

1505

REVIEWS

J. Rebek, Jr. et al.

Figure 30. Stabilization of reactive species through encapsulation: a) for- mation of unique tennis ball shaped hydrophobic dimers 72 and 73 within a metal-based supramolecular capsule 41 a ; b) encapsulation and stabiliza- tion of highly reactive cyclic trisilanols 74 a ± c, proposed to be intermedi- ates in the sol±gel polycondensation process; c)formation of a water- sensitive phosphane ± acetone adduct 75 within a tetrahedral metal ± ligand cluster 46 in the aqueous phase; d) the cylindrical capsule 18 is able to prevent reactions of the shape-complementary guest benzoyl peroxide (76) even at elevated temperatures for prolonged periods of time, even though the capsule structure is maintained entirely through hydrogen bonds.

clic trimer had not been observed. Evidence for encapsulated cyclic trimer 74 is provided by NMR spectroscopy, mass spectrometry, and, for 74c, X-ray crystallography.[111] In all cases, the trimer is formed as the C3-symmetric all-cis isomer. The reactivity of 74 is greatly attenuated by encapsulation. No change in the complex is observed for over one month in neutral aqueous solution; the guest also survives the acidic conditions (pH ` 1) required for the isolation of the com- plexes. Again, the guest is too large for formation outside the capsule; instead, the polycondensation process that traps this reactive intermediate must take place within the confines of the capsule walls.

The tetrahedral metal clusters of Raymond and co-work- ers[101] offer a radical approach for the stabilization of reactive intermediates. The treatment of a solution of Ga ± catecholate based capsule 46 (Figure 15) in D2O with PEt3 resulted in the observation of new signals for encapsulated guest in the 1H and 31P NMR spectra. These signals did not correlate with the expected encapsulation of PEt3 , but instead can be attributed to the encapsulation of the cationic phosphane ± acetone adduct [Me2C(OH)PEt3] (75) that arises as a consequence of the presence of adventitious acetone remaining from the synthesis of capsule 46 (Figure 30 c).[146] This adduct has been previously synthesized under anhydrous conditions, but decomposes rapidly in aqueous solution as a result of the low concentration of acetone. It is likely that 75 forms upon entry of protonated phosphane into a cavity already contain-

ing residual acetone. To confirm the structure of the encapsulated species, [Me2C(OH)PEt3]Br was prepared un- der anhydrous conditions and added to the capsule in CD3OD. The resulting 1H and 31P NMR spectra agreed with those obtained previously in D2O. Mass spectrometric studies of the methanolic solution provided further evidence for the com- position of the encapsulated species.

Even capsules constructed through relatively weak hydro- gen bonds can act to stabilize reactive species through encapsulation. Benzoyl peroxide (76) readily undergoes homolytic bond cleavage at room temperature to give reactive radical species that are commonly used for the initiation of radical chain reactions or the oxidation of various substrates. The size and shape of 76 make it an excellent guest for capsule 18 (Figure 30 d).[147] A variety of agents normally reactive to 76 at room temperature undergo no detectable reaction in the presence of the preformed encapsulation complex of 76 and 18 during prolonged heating at 70VC. The addition of a small amount of DMF disrupts the hydrogen bonds responsible for encapsulation and results in immediate reaction of the reporter molecules with 76. Likewise, the addition of a competing guest molecule that displaces 76 from the capsule also results in the release of the reactive species and the onset of oxidation or chain reactions. Studies on complexes of other guests with 18 indicate that the partial opening of the capsule walls is probably responsible for the rapid exchange of small guests, but little is known about the exchange of large rod- shaped guests. Perhaps the complete dissociation of the complex–a rare event–is required. The surprising stability of this complex highlights the subtle effect of shape-selective

1506

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

host ± guest interactions on guest-binding processes.

6. Summary and Outlook

and -exchange

The final paragraph of a review asks the writer to predict the future–or worse, to tip his hand about the direction his own research will take. Naturally, we are hesitant to do either, but we confess an interest in the construction of synthetic systems that possess nonlinear properties such as autocatalysis and chemical amplification. These characteristics are integral properties of living systems, and they give rise to desirable behaviors such as increased sensitivity, responsiveness, and self-replication. Other questions that remain to be answered lie in the realm of supramolecular mechanisms. In simple encapsulation complexes, an understanding of paired receptor and guest movement during binding and release events is attainable, whereas in the biological realm of complex receptor ± ligand systems the flexibility and intricate move- ment of a receptor and guest during a binding event are difficult to study, and are often overlooked.[148] The study of simple systems does not limit the details of questions that may be asked, instead, exploring well-defined systems can allow for the understanding of intricacies that otherwise may not even be considered.

We emphasize that reversible encapsulation is not intended as a model of anything; it provides the current outlet for our curiosity about the nature of intimate molecular relationships

Molecular Capsules

REVIEWS

and their manipulation. Yet we cannot deny that the compel- ling nature of encapsulation has its roots in biology. There, compartmentalization provides ways to separate incompat- ible reagents and environments (endosomes and mitochon- dria isolate media of widely differing pH values). We believe that reversible encapsulation can provide a probe operating at the boundary of chemistry with biology, the most intriguing of which is how the former gave rise to the latter–the TMWest Coast∫ approach to chemical biology. Given the activity at The Skaggs Institute,[15, 149, 150] the Scripps Oceanographic Institute,[151]theUniversityofCaliforniaatSanDiego,[152] and the Salk Institute,[9] what better location is there to explore this question than La Jolla?

This review is dedicated to Ivar Ugi, the architect of MolecularDiversity.TheauthorsthankDr.LubomirSebo for expert assistance with graphical design and the Skaggs Institute for Chemical Biology for funding. F.H. thanks the NSERC for a graduate fellowship.

Received: June 25, 2001 [A480]

[1] This is alleged by no less of an authority than Woody Harrelson in the movie Indecent Proposal (USA, 1993). For the original Khan quote, see The Encyclopedia of Twentieth-Century Architecture (Ed.: R. StephenSennott),FitzroyDearborn,Chicago,2002.

[27]J.Canceill,L.Lacombe,A.Collet,J.Am.Chem.Soc.1986,108,4230. [28]J.S.Lindsey,New.J.Chem.1991,15,153. [29]G.M.Whitesides,J.P.Mathias,C.T.Seto,Science1991,254,1312. [30]D.Philp,J.F.Stoddart,Syn.Lett.1991,445. [31]J.M.Lehn,Angew.Chem.1990,102,1347;Angew.Chem.Int.Ed.

Engl. 1990, 29, 1304. [32]R.C.Dunbar,T.B.McMahon,Science1998,279,194. [33]D.A.Evans,J.S.Evans,J.Org.Chem.1998,63,8027. [34]K.T.Chapman,W.C.Still,J.Am.Chem.Soc.1989,111,3075. [35]A.E.Rowan,J.A.A.W.Elemans,R.J.M.Nolte,Acc.Chem.Res.

1999, 32, 995.
[36]J.Rebek,Jr.,Acc.Chem.Res.1999,32,278. [37]R.Wyler,J.deMendoza,J.Rebek,Jr.,Angew.Chem.1993,105,

1820; Angew. Chem. Int. Ed. Engl. 1993, 32, 1699. [38]N.Branda,R.Wyler,J.Rebek,Jr.,Science1994,263,1267. [39]N.Branda,R.M.Grotzfeld,C.Valdes,J.Rebek,Jr.,J.Am.Chem.

Soc. 1995, 117, 85. [40]R.S.Meissner,J.Rebek,Jr.,J.deMendoza,Science1995,270,1485. [41]J.Kang,J.Rebek,Jr.,Nature1996,382,239. [42]J.M.Rivera,T.Martin,J.Rebek,Jr.,J.Am.Chem.Soc.1998,120,

819. [43]R.M.Grotzfeld,N.Branda,J.Rebek,Jr.,Science1996,271,487. [44]T.Szabo,B.M.O×Leary,J.Rebek,Jr.,Angew.Chem.1998,110,

3606; Angew. Chem. Int. Ed. 1998, 37, 3410.
[45] C. B. Aakerˆy, A. M. Beatty, B. A. Helfrich, Angew. Chem. 2001, 113,

3340; Angew. Chem. Int. Ed. 2001, 40, 3240.
[46] T. Martin, U. Obst, J. Rebek, Jr., Science 1998, 281, 1842.
[47] F. Hof, C. Nuckolls, S. L. Craig, T. Martin, J. Rebek, Jr., J. Am. Chem.

Soc. 2000, 122, 10 991. [48]J.Rebek,Jr.,Chem.Commun.2000,637. [49]K.D.Shimizu,J.Rebek,Jr.,Proc.Natl.Acad.Sci.USA1995,92,

[2]J.Rebek,Jr.,Acc.Chem.Res.1984,17,258. [3]S.C.Zimmerman,B.F.Duerr,J.Org.Chem.1992,57,2215. 12403.

[4]S.C.Zimmerman,P.Schmitt,J.Am.Chem.Soc.1995,117,10769. [5]S.Anderson,H.L.Anderson,J.K.M.Sanders,Acc.Chem.Res.

1993, 26, 469. [6]R.Hoss,F.Vˆgtle,Angew.Chem.1994,106,389;Angew.Chem.Int.

Ed.Engl.1994,33,375. [7]TemplatedOrganicSynthesis(Eds.:F.Diederich,P.J.Stang),Wiley-

VCH, Weinheim, 1999. [8]G.vonKiedrowski,Angew.Chem.1986,98,932;Angew.Chem.Int.

Ed. Engl. 1986, 932. [9]L.E.Orgel,Nature1992,358,203.

[10]S.Pitsch,R.Krishnamurthy,M.Bolli,S.Wendeborn,A.Holzner,M. Minton,C.Lesueur,I.Scholonvogt,B.Jaun,A.Eschenmoser,Helv. Chim. Acta 1995, 78, 1621.

[11]T.Tjivikua,P.Ballester,J.Rebek,Jr.,J.Am.Chem.Soc.1990,112, 1249.

[12]T.Li,K.C.Nicolaou,Nature1994,369,218.

  1. [13]  A. Terfort, G. von Kiedrowski, Angew. Chem. 1992, 104, 626; Angew.Chem. Int. Ed. Engl. 1992, 31, 654.
  2. [14]  B. Wang, I. O. Sutherland, Chem. Commun. 1997, 1495.

[15]D.H.Lee,J.R.Granja,J.A.Martinez,K.Severin,M.R.Ghadiri,

Nature 1996, 382, 525.
[16] S. Yao, I. Ghosh, R. Zutshi, J. Chmielewski, J. Am. Chem. Soc. 1997,

119, 10 559.
[17] T. R. Kelly, C. Zhao, G. J. Bridger, J. Am. Chem. Soc. 1989, 111, 3744. [18]D.Philp,J.F.Stoddart,Angew.Chem.1996,108,1242;Angew.Chem.

Int. Ed. Engl. 1996, 35, 1155. [19]C.J.Walter,H.L.Anderson,J.K.M.Sanders,Chem.Commun.

1993, 459.
[20]J.Rebek,Jr.,Science1987,235,1478. [21]J.C.Adrian,Jr.,C.S.Wilcox,J.Am.Chem.Soc.1989,111,8055. [22]S.Zimmerman,W.Wu,J.Am.Chem.Soc.1989,111,8054. [23]S.P.Brown,T.Schaller,U.P.Seelbach,F.Koziol,C.Ochsenfeld,F.-

G.Kl‰rner,H.W.Spiess,Angew.Chem.2001,113,740;Angew.

Chem. Int. Ed. 2001, 40, 717. [24]P.G.E.Sanderson,J.D.Kilburn,W.C.Still,J.Am.Chem.Soc.1989,

111, 8314. [25]R.P.Dixon,S.J.Gelb,A.D.Hamilton,J.Am.Chem.Soc.1992,114,

365. [26]D.J.Cram,H.-J.Choi,J.A.Bryant,C.B.Knobler,J.Am.Chem.Soc.

1992, 114, 7748.
Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508

[50]J.Scheerder,J.P.M.vanDuynhoven,J.F.J.Engbersen,D.N. Reinhoudt,Angew.Chem.1996,108,1172;Angew.Chem.Int.Ed. Engl. 1996, 35, 1090.

[51]O.Mogck,V.Bˆhmer,W.Vogt,Tetrahedron1996,52,8489. [52]R.K.Castellano,J.Rebek,Jr.,J.Am.Chem.Soc.1998,120,3657. [53]M.O.Vysotsky,I.Thondorf,V.Bˆhmer,Angew.Chem.2000,112,

1309; Angew. Chem. Int. Ed. 2000, 39, 1264. [54]R.K.Castellano,B.H.Kim,J.Rebek,Jr.,J.Am.Chem.Soc.1997,

119, 12671. [55]R.K.Castellano,C.Nuckolls,J.Rebek,Jr.,J.Am.Chem.Soc.1999,

121,11156. [56]J.J.Gonzalez,R.Ferdani,E.Albertini,J.M.Blasco,A.Arduini,A.

Pochini, P. Prados, J. de Mendoza, Chem. Eur. J. 2000, 6, 73. [57]L.R.MacGillivray,J.L.Atwood,Nature1997,389,469. [58]T.Gerkensmeier,W.Iwanek,C.Agena,R.Froelich,S.Kotila,C.

Naether,J.Mattay,Eur.J.Org.Chem.1999,2257.
[59] Y. Aoyama, Y. Tanaka, H. Toi, H. Ogoshi, J. Am. Chem. Soc. 1988,

110, 634.
[60] A. Shivanyuk, J. Rebek, Jr., Proc. Natl. Acad. Sci. USA 2001, 98, 7662. [61]K.Murayama,K.Akoi,Chem.Commun.1998,607.
[62] A. Shivanyuk, K. Rissanen, E. Kolehmainen, Chem. Commun. 2000,

1107.
[63] K. N. Rose, L. J. Barbour, G. W. Orr, J. L. Atwood, Chem. Commun.

1998, 407. [64]P.Soncini,S.Bonsignore,E.Dalcanale,F.Ugozzoli,J.Org.Chem.

1992, 57, 4608. [65]M.Vincenti,C.Minero,E.Pelizzetti,A.Secchi,E.Dalcanale,Pure

Appl. Chem. 1995, 67, 1075. [66]T.Heinz,D.M.Rudkevich,J.Rebek,Jr.,Nature1998,394,764. [67]R.G.Chapman,J.C.Sherman,J.Am.Chem.Soc.1998,120,9818. [68]K.Kobayashi,T.Shirasaka,E.Horn,N.Furukawa,K.Yamaguchi,S.

Sakamoto,Chem.Commun.2000,41. [69]A.Shivanyuk,E.F.Paulus,V.Bˆhmer,Angew.Chem.1999,111,

3091; Angew. Chem. Int. Ed. 1999, 38, 2906. [70]A.Luetzen,A.R.Renslo,C.A.Schalley,B.M.O×Leary,J.Re-

bek, Jr., J. Am. Chem. Soc. 1999, 121, 7455. [71]B.J.Holliday,C.A.Mirkin,Angew.Chem.1993,113,2076;Angew.

Chem. Int. Ed. 2001, 40, 2022. [72]N.Cuminetti,M.H.K.Ebbing,P.Prados,J.deMendoza,E.

Dalcanale, Tetrahedron Lett. 2001, 42, 527.

1507

REVIEWS

J. Rebek, Jr. et al.

[73]P.Jacopozzi,E.Dalcanale,Angew.Chem.1997,109,665;Angew. Chem. Int. Ed. Engl. 1997, 36, 613.

[74]E.Dalcanale,J.Am.Chem.Soc.2001,123,7539. [75]Z.Zhong,A.Ikeda,M.Ayabe,S.Shinkai,S.Sakamoto,K.

Yamaguchi,J.Org.Chem.2001,66,1002. [76]A.Ikeda,M.Yoshimura,H.Udzu,C.Fukuhara,S.Shinkai,J.Am.

Chem. Soc. 1999, 121, 4296. [77]O.D.Fox,N.K.Dalley,R.G.Harrison,J.Am.Chem.Soc.1998,120,

7111. [78]O.D.Fox,N.K.Dalley,R.G.Harrison,Inorg.Chem.1999,38,5860. [79]O.D.Fox,N.K.Dalley,R.G.Harrison,Inorg.Chem.2000,39,620. [80]O.D.Fox,J.F.Y.Leung,J.M.Hunter,N.K.Dalley,R.G.Harrison,

Inorg.Chem.2000,39,783. [81]G.W.Orr,L.J.Barbour,J.L.Atwood,Science1999,285,1049. [82]M.Fujita,K.Umemoto,M.Yoshimura,N.Fujita,T.Kusukawa,K.

Biradha, Chem. Commun. 2001, 509. [83]M.Fujita,S.Nagao,K.Ogura,J.Am.Chem.Soc.1995,117,1649. [84]M.Fujita,D.Oguro,M.Miyazawa,H.Oka,K.Yamaguchi,K.Ogura,

Nature 1995, 378, 469. [85]F.Ibukuro,T.Kusukawa,M.Fujita,J.Am.Chem.Soc.1998,120,

8561. [86]T.Kusukawa,M.Fujita,Angew.Chem.1998,110,3327;Angew.

Chem. Int. Ed. 1998, 37, 3142. [87]N.Takeda,K.Umemoto,K.Yamaguchi,M.Fujita,Nature1999,398,

794. [88]K.Umemoto,H.Tsukui,T.Kusukawa,K.Biradha,M.Fujita,

Angew.Chem.2001,113,2690;Angew.Chem.Int.Ed.2001,40,2620. [89]S.-Y.Yu,T.Kusukawa,K.Biradha,M.Fujita,J.Am.Chem.Soc.

2000,122,2665.

  1. [90]  K. Umemoto, K. Yamaguchi, M. Fujita, J. Am. Chem. Soc. 2000, 122,7150.
  2. [91]  R. W. Saalfrank, A. Stark, M. Bremer, H. U. Hummel, Angew. Chem.1990, 102, 292; Angew. Chem. Int. Ed. Engl. 1990, 29, 311.

[92]R.W.Saalfrank,B.Hoerner,D.Stalke,J.Salbeck,Angew.Chem.

1993,105,1223;Angew.Chem.Int.Ed.Engl.1993,32,1179. [93]R.W.Saalfrank,R.Burak,A.Breit,D.Stalke,R.Herbst-Irmer,J. Daub, M. Porsch, E. Bill, M. Muether, A. X. Trautwein, Angew. Chem.1994,106,1697;Angew.Chem.Int.Ed.Engl.1994,33,1621. [94] S. Mann, G. Huttner, L. Zsolnai, K. Heinze, Angew. Chem. 1996, 108,

[111]T.Kusukawa,M.Yoshizawa,M.Fujita,Angew.Chem.2001,113, 1931; Angew. Chem. Int. Ed. Engl. 2001, 40, 1879.

[112]B.M.O×Leary,R.M.Grotzfeld,J.Rebek,Jr.,J.Am.Chem.Soc. 1997,119,11701.

[113]R.G.Chapman,J.C.Sherman,J.Org.Chem.2000,65,513. [114]J.Santamaria,T.Martin,G.Hilmersson,S.L.Craig,J.Rebek,Jr.,

Proc. Natl. Acad. Sci. USA 1999, 96, 8344. [115]L.J.Prins,F.DeJong,P.Timmerman,D.N.Reinhoudt,Nature2000,

408, 181. [116]J.M.Ribo,J.Crusats,F.Sagues,J.Claret,R.Rubires,Science2001,

292,2063. [117]Y.Tokunaga,J.Rebek,Jr.,J.Am.Chem.Soc.1998,120,66. [118]J.M.Rivera,T.Martin,J.Rebek,Jr.,Science1998,279,1021. [119]J.M.Rivera,T.Martin,J.Rebek,Jr.,J.Am.Chem.Soc.2001,123,

5213. [120]J.M.Rivera,S.L.Craig,T.Martin,J.Rebek,Jr.,Angew.Chem.2000,

112,2214;Angew.Chem.Int.Ed.2000,39,2130. [121]A.Ikeda,H.Udzu,Z.Zhong,S.Shinkai,S.Sakamoto,K.

Yamaguchi, J. Am. Chem. Soc. 2001, 123, 3872. [122]C.Nuckolls,F.Hof,T.Martin,J.Rebek,Jr.,J.Am.Chem.Soc.1999,

121, 10281. [123]A.J.Terpin,M.Ziegler,D.W.Johnson,K.N.Raymond,Angew.

2983;Angew.Chem.Int.Ed.Engl.1996,35,2808.
[95]C.M.Hartshorn,P.J.Steel,Chem.Commun.1997,541. Lovinger,J.Rebek,Jr.,Angew.Chem.1999,111,2764;Angew.

[96]J.S.Fleming,K.L.V.Mann,C.-A.Carraz,E.Psillakis,J.C.Jeffery, J.A.McCleverty,M.D.Ward,Angew.Chem.1998,110,1315; Angew. Chem. Int. Ed. 1998, 37, 1279.

[97]S.S.Y.Chui,S.M.F.Lo,J.P.H.Charmant,A.G.Orpen,I.D. Williams,Science1999,283,1148.

[98]F.A.Cotton,L.M.Caniels,C.Lin,C.A.Murillo,Chem.Commun. 1999,841.

[99]R.L.Paul,S.M.Couchman,J.C.Jeffery,J.A.McCleverty,Z.R. Reeves,M.D.Ward,Dalton2000,845.

[100]D.L.Caulder,R.E.Powers,T.N.Parac,K.N.Raymond,Angew. Chem. 1998, 110, 1940; Angew. Chem. Int. Ed. Engl. 1998, 37, 1840. [101]D.L.Caulder,C.Br ̧ckner,R.E.Powers,S.Kˆnig,T.N.Parac,J.A.

Leary, K. N. Raymond, J. Am. Chem. Soc. 2001, 123, 8923. [102]T.N.Parac,D.L.Caulder,K.N.Raymond,J.Am.Chem.Soc.1998,

120, 8003. [103]T.N.Parac,M.Scherer,K.N.Raymond,Angew.Chem.2000,112,

Chem.Int.Ed.1999,38,2603. [136]R.K.Castellano,R.Clark,S.L.Craig,C.Nuckolls,J.Rebek,Jr.,

Proc. Natl. Acad. Sci. USA 2000, 97, 12418. [137]J.Kang,J.Rebek,Jr.,Nature1997,385,50. [138]J.Kang,G.Hilmersson,J.Santamaria,J.Rebek,Jr.,J.Am.Chem.

Soc.1998,120,3650. [139]J.Kang,J.Santamaria,G.Hilmersson,J.Rebek,Jr.,J.Am.Chem.

Soc.1998,120,7389. [140]H.Ito,T.Kusukawa,M.Fujita,Chem.Lett.2000,598. [141]J.Chen,S.Kˆrner,S.L.Craig,D.Rudkevich,J.Rebek,Jr.,Nature

2002, 415, 385. [142]D.J.Cram,M.E.Tanner,R.Thomas,Angew.Chem.1991,103,1048;

Angew. Chem. Int. Ed. Engl. 1991, 30, 1024. [143]T.Heinz,D.M.Rudkevich,J.Rebek,Jr.,Angew.Chem.1999,111,

1206; Angew. Chem. Int. Ed. 1999, 38, 1136. [144]T.Kusukawa,M.Fujita,J.Am.Chem.Soc.1999,121,1397. [145]M.Yoshizawa,T.Kusukawa,M.Fujita,K.Yamaguchi,J.Am.Chem.

1288; Angew. Chem. Int. Ed. 2000, 39, 1239. [104]D.W.Johnson,K.N.Raymond,Inorg.Chem.2001,40,5157. Soc.2000,122,6311.

[105]M.Scherer,D.L.Caulder,D.W.Johnson,K.N.Raymond,Angew. Chem. 1999, 111, 1690; Angew. Chem. Int. Ed. 1999, 38, 1588.

  1. [106]  X. Sun, D. W. Johnson, D. L. Caulder, K. N. Raymond, E. H. Wong, J.Am. Chem. Soc. 2001, 123, 2752.
  2. [107]  L. Frish, S. E. Matthews, V. Bˆhmer, Y. Cohen, Perkin Trans. 2 1999,669.

[108]G.Hilmersson,J.Rebek,Jr.,Magn.Reson.Chem.1998,36,663. [109]O.Mogck,M.Pons,V.Bˆhmer,W.Vogt,J.Am.Chem.Soc.1997,119,

5706. [110]V.Bˆhmer,O.Mogck,M.Pons,E.F.Paulus,NATOASISer.Ser.C

1999, 526, 45.

[146]M.Ziegler,J.L.Brumaghim,K.N.Raymond,Angew.Chem.2000, 112, 4285; Angew. Chem. Int. Ed. 2000, 39, 4119.

[124] [125]

Chem. 2001, 113, 161; Angew. Chem. Int. Ed. 2001, 40, 157.
J.-M. Lehn, Chem. Eur. J. 1999, 5, 2455.
G. R. L. Cousins, S.-A. Poulsen, J. K. M. Sanders, Curr. Opin. Chem. Biol.2000,4,270.

[126]V.Goral,M.I.Nelen,A.V.Eliseev,J.-M.Lehn,Proc.Natl.Acad.Sci. USA2001,98,1347.

[127]J.-M.Lehn,A.V.Eliseev,Science2001,291,2331. [128]F.Hof,C.Nuckolls,J.Rebek,Jr.,J.Am.Chem.Soc.2000,122,4251. [129]C.A.Schalley,Int.J.MassSpectrom.2000,194,11. [130]S.Sakamoto,M.Yoshizawa,T.Kusukawa,M.Fujita,K.Yamaguchi,

Org. Lett. 2001, 3, 1601. [131]S.Hiraoka,M.Fujita,J.Am.Chem.Soc.1999,121,10239. [132]S.Hiraoka,Y.Kubota,M.Fujita,Chem.Commun.2000,1509. [133]R.K.Castellano,S.L.Craig,C.Nuckolls,J.Rebek,Jr.,J.Am.Chem.

Soc. 2000, 122, 7876. [134]R.K.Castellano,D.M.Rudkevich,J.Rebek,Jr.,Proc.Natl.Acad.

Sci. USA 1997, 94, 7132. [135]R.K.Castellano,C.Nuckolls,S.H.Eichhorn,M.R.Wood,A.J.

S. K. Kˆrner, F. C. Tucci, D. M. Rudkevich, T. Heinz, J. Rebek, Jr., Chem. Eur. J. 2000, 6, 187.
A. C. Anderson, R. H. O×Neil, T. S. Surti, R. M. Stroud, Chem. Biol. 2001, 8, 445.

[147]
[148] [149]M.Bolli,R.Micura,A.Eschenmoser,Chem.Biol.1997,4,309.

[150] [151]

[152]

M. C. Wright, G. F. Joyce, Science 1997, 276, 614.
G. Arrhenius, K. K. Baldridge, S. Richards-Gross, J. S. Siegel, J. Org. Chem.1997,62,5522.
H. J. Cleaves, S. L. Miller, Proc. Natl. Acad. Sci. USA 1998, 95, 7260.

1508

Angew. Chem. Int. Ed. 2002, 41, 1488 ± 1508