Anion-Templated Self-Assembly: Organic Compounds Part 1 (Nanotechnology)

INTRODUCTION

The field of self-assembly is an increasingly attractive area of supramolecular chemistry. With many principles derived from biological systems, it often enables the synthesis of large, complex structures that would be far too demanding using conventional covalent tech-niques.[1-3] Utilizing weak, reversible, noncovalent interactions, self-assembly can, through rational design of simple ligands, help minimize the amount of information inherent in the building blocks for the system and allow for error checking and self-correction.[4-6] To date, the range of noncovalent interactions used in self-assembly processes includes hydrogen bonding, metal coordination, hydrophobic, and p-p donor-acceptor interactions.[7-18] The range of structures prepared via templated self-assembly continues to become increasingly diverse and intricate.[7-18]

Interest in the binding and recognition of anionic guest species has increasingly grown in the last decade or so, such that the field is now considered an important area of supramolecular chemistry. Anions often play key roles in biological processes; many enzyme substrates and cofactors are anionic and DNA itself is a polyanion. The fields of medicine and catalysis also help illustrate the diverse areas in which anions play key roles. Some anions are of environmental concern (e.g., there is a need to sense and remove nitrate and phosphate pollutants from natural waterways). In the past two decades, progress in the binding and recognition of anions has advanced considerably, and there are now many reviews dedicated entirely to this specific area of supramolecular chemistry.[19-22]

In light of this, it is surprising that investigations into the use of anions as templates in the self-assembly of supramolecular architectures has, until recently, been limited. Reasons for this may stem from the anion’s small charge-to-radius ratio (more diffuse nature), pH sensitivity, and high solvation energy. The large geometrical diversity found in anions indicates the possibilities for predefining a wide range of supramolecular structures.

At the time of writing, we are unaware of any reviews dedicated solely to anion-templated self-assembly and are thus led to believe that this topic offers the first reviews of this topic.a Readers interested in anion-templated self-assembly of inorganic based frameworks are directed to the article ”Anion-Templated Self-Assembly: Inorganic Compounds.” This article focuses solely on the anion-templated self-assembly of organic compounds.

ANION-TEMPLATED SELF-ASSEMBLY OF ORGANIC FRAMEWORKS

The range of organic supramolecules prepared via anion-templated self-assembly is highly varied. Examples given in this review include solid-state polymers, helicates, pseudorotaxanes, and rotaxanes. Although nuclear magnetic resonance (NMR) and mass spectrometry data provide solution evidence of assembly in many examples, X-ray crystallography remains an invaluable tool for confirmation of the role of anions in many of the early reports.

The first example of anion templation in organic systems is in the rational design of organic solid-state structures reported by Geib et al.[24] in 1991. In contrast to traditional endo hydrogen bonding sites that form discrete 1:1 complexes with complementary receptors, they illustrate that by utilizing an exo configuration, the same components can be made to assemble into an alternating polymeric structure.

By simple protonation of the pyridine ring in 2,6-dibutyramidopyridine, repulsive interactions force rotation about the pyridine-amide bond, leading to the formation of two intramolecular hydrogen bonds to the carbonyl oxygen. Therefore the amide-NH groups are outwardly (or exo) directed and thus available for in-termolecular hydrogen bonding. Addition of bis(4-nitro-phenyl)hydrogenphosphate not only results in protonation but leads to the self-assembly of a polymeric alternating ribbon structure. The diaryl hydrogen phosphate plays two key roles in this assembly: protonation of the pyridine ring and provision of an anionic hydrogen bond acceptor of suitable geometry. The ribbon arrangement has hydrogen bonds in its core, with the anionic and cation-ic components segregated onto opposite sides of the structure in parallel alignment to each other (Fig. 1). A further structure was assembled using diphenyl hydrogen phosphate, indicating the generality of this assembling strategy.

Fig. 1 Ribbon arrangement of Hamilton’s self-assembled solid-state hydrogen-bonded structure.

The first example of an anion-templated helicate was reported by Sanchez-Quesada et al.[25] in 1996. A tetra-guanidinium strand 1 was prepared, in which the spacer unit is too short for it to be able to wrap around one sulphate anion (Fig. 2). Consequently, two strands are forced to self-assemble into a double helicate structure, the handedness of which is imposed by the chiral receptor. Evidence for the helical nature of the anion-templated structure is provided by ROESY (Rotational nuclear Overhauser Effect SpectroscopY) NMR techniques. In addition and in contrast to many helicates assembled by coordination chemistry, these organic heli-cates are overall neutral.

In 1996, Sessler et al.[26] reported on the self-assembly of polypyrrolic macrocycles. Based on the anion binding properties of sapphyrins and calix[4]pyrroles, zwitterionic carboxylate-appended structures 2 and 3 (Fig. 3) were synthesized. The ability to bind intramolecularly is absent because of the choice of spacer; with the sapphyrin compound methylene, carboxylate groups are positioned at either R2 or R4, with the other R groups being hydrogen or methyl groups, and on the calixpyrrole, R is a pro-pylene carboxylate.

Fig. 2 Mendoza’s tetraguanidinium strand self-assembles into a double helicate structure.

It was found that these structures assemble into dimers; the carboxylate group of one molecule is chelated by the pyrrole core of another (Fig. 4). The tail of this second molecule is subsequently bound by the first molecule.

Protonated sapphyrins are known to have a high affinity for fluoride anions. Addition of fluoride ions is shown to inhibit the dimer assembly, thus providing further evidence for the assembly process. This example demonstrates how known anion binding sites can be exploited in the self-assembly of discrete supramolecular species.

Sessler et al.[27] have also reported on the possible templating effect of anionic nitrates in the synthesis of an oligopyrrolic macrocycle. An unusual example of large polynuclear anions as templates in organic synthesis is provided by Kim et al.,[28] who utilized the [CaCl3-(DMAc)3]~ (DMAc=dimethyl acetamide) pseudo-octahedral anion in a directing role in the preparation of cyclic aromatic amides.

The use of anions in the synthesis of [I4]imidazolio-phanes was reported by Alcalde et al.[29,30] in 1999. Based on a [3+1] convergent cyclization (Scheme 1), yields are significantly enhanced by the presence of certain anions (e.g., 42% yield in the absence of anions, to 83% for chloride and 88% for bromide). Templating is postulated to be because of the formation of an intermediate in which C-H.. .Cl_ hydrogen bonds force a conformation that favors subsequent cyclization.

In recent years, interest in the preparation of mechanically interlocked supramolecules, such as catenanes and rotaxanes, has grown immensely. Rotaxanes are systems where a threadlike molecule is encircled by a macrocycle. Large stopper groups prevent the macrocycle from slipping off the ends of the thread. These molecules present a great challenge to the synthetic chemist because of their unusual nature and the possibilities for unusual molecular properties.[31-37] An extension of rotaxane work is the ever-expanding field of molecular shuttles and machines.[31-37] To date, the majority of rotaxane preparations have involved the use of hydrogen bonds, metal cation coordination chemistry, and p-p donor-acceptor interactions to mediate the assembly.[38-42] More recently, some research groups have started to look into the possibility of using anions to template the formation of these interlocked, complex structures.

Scheme 1 Alcalde’s convergent synthesis of [I4]imidazolio-phanes.

Fig. 3 Carboxylate-appended sapphyrins and calix[4]pyrroles.

In 1997, Fyfe et al.[43] reported what they describe as ”anion-assisted self-assembly” of polypseudorotaxanes. Pseudorotaxanes are supramolecular complexes in which one molecule is threaded through another, but the absence of stopper groups means they can subsequently dissociate. These equilibrium complexes are important as they not only provide a great deal of information on the assembling motif but are precursors to both rotaxanes and catenanes. Earlier work by Montalti and Prodi[44] and Ashton et al.[45,46] established a strategy for pseu-dorotaxane formation constructed from macrocyclic polyethers and dibenzylammonium ions. Favorable [N+-H.. .O] and [C-H.. .O] hydrogen bonds and p-p stacking interactions help thread the ammonium ions through the macrocycles. Extension of this work to higher-order pseudorotaxanes hinted at the organizational role of an encapsulated PFT ion within the interior of the supramolecule.

With tri-p-phenylene[51]crown-15, there lies the possibility of incorporating three ammonium ions into the cavity, thus forming a [4]pseudorotaxane (Fig. 5). Each ammonium ion interacts independently with a single polyether loop by the expected [N+-H.. .O] and [C-H.. .O] interactions. Crystal structure evidence reveals that a single PFT anion is located centrally and partially encapsulated within a cleft generated by the saddlelike conformation of the complex. A series of [C-H.. .F] hydrogen bonds to the hydroquinone rings on the polyether macrocycle and to the ammonium benzylic methylene groups stabilizes and imposes order on the PFT ion.

Fig. 4 Side view of the noncovalent calix[4]pyrrole dimer where R=(CH2)3CO2H.

Fig. 5 Stoddart’s [4]pseudorotaxane.

Fig. 6 Stoddart’s [5]pseudorotaxane.

The larger [5]pseudorotaxane formed between four ammonium threads and tetrakis-p-phenylene[68]crown-20 also features an ordered PFT ion (Fig. 6). In this case, the anion is completely encapsulated by the four tetrahedrally arranged ammonium ions and by the four hydroquinone rings of the macrocycle. Order is imposed on the octahedral anion by a series of [C-H.. .F] hydrogen bonds involving the methylene hydrogen atoms of the threads and the hydroquinone protons of the poly-ether ring.

Stoddart et al. conclude that the PFT anion ”programs” the geometry of both complexes. However, although the anion undoubtedly plays a role in the assembly of these complexes and the final geometry of the system, it is worth noting that secondary ammonium threads have been shown to assemble with polyether chains, especially dibenzo[24]crown[8], where the counter-anion is found to play no role in the assembly.

Fig. 7 Schematic of Smith’s ion pair binding rotaxane indicating the predominant co-conformation of axle and wheel orientation in CDCl3. Tr=Trityl.

The first example of anion-templated synthesis of ro-taxanes was reported by Hubner et al.[47] in 1999. Relying on the anion recognition properties of macrocyclic lac-tams, the synthesis proceeds via a ”supramolecular nu-cleophile” or ”wheeled phenoxide.” The hydrogen bond-donating ability of the lactam wheel is used to complex an organic anion, which can then serve as a nucleophile in an Sn2 reaction. The reaction of p-tritylphenolate bound inside the lactam wheel with the suitable axle component gives the [2]rotaxane (Scheme 2) in a remarkably high yield of 95%. It is worth noting that the rotaxane product is neutral; the anion template is ”used up” in the synthesis.[48]

Further reports have demonstrated the versatility of this so-called ”trapping” methodology using a wide variety of axle building blocks. Rotaxanes with carbonate, acetal, and ester centerpieces have been prepared in varying yields.[49] Attempts to extend the synthesis to proceed via carboxylate anions have proven problematic and demonstrate that phenolate formation is essential in the preparation of these rotaxanes.

Scheme 2 Vogtle’s anion-templated rotaxane synthesis.

Smith et al. have tried to exploit the trapping technique developed by Vogtle in the preparation of ion pair binding [2]rotaxanes. Earlier ion pair binding macrocycles developed by Smith et al. incorporate an anion binding isophthalamide cleft that bridges a cation binding crown region and a close structural analogue has now been used to form the wheel of new rotaxanes.[50,51] In these macrocyclic systems, it was found that not only could an alkali metal cation and an anion be bound simultaneously, but that the anion affinity is enhanced by the presence of suitable cations. It has been shown that the new host is capable of binding the potassium salt of 4-tritylphenolate and that subsequent reaction of the ”wheeled phenolate” with isophthaloyl dichloride yields a thermodynamically stable [2]rotaxane in 20% yield (Fig. 7).

Evidence for rotaxane formation is provided by fast atom bombardment (FAB) mass spectrometry and 1H NMR assignments carried out by a combination of homo-nuclear correlation spectroscopy (COSY) and ROESY methods. ROESY evidence also suggests that the presence of K+ cations freezes out a single co-conformation. In the absence of K+ cations, broad signals that sharpen on heating, indicating a number of axle-wheel orientations, are observed at room temperature. This cation-dependent behavior hints at future work whereby molecular motion could be controlled. The binding of chloride has been found to have an effect on the dynamic behavior of the rotaxane, although the combined effect of K+/CP binding produces the most rigid structure.[52,53]

Scheme 3 Schalley’s ”stoppering” rotaxane synthesis using an anion template.

Fig. 8 Wisner et. al.[59] have developed a powerful and versatile anion-templated assembling motif for interlocked molecules.

Since this initial report, Deetz et al.[53] have embarked on a systematic study of their ion pair binding rotaxanes including rotaxanes with larger anion binding cavities and longer acetal-based threads, which could potentially be used in molecular shuttle devices.