Supramolecular Polymers and Assemblies. Andreas Winter
can still utilize that numerous features of the living ionic or the radical polymerization explain the cooperative supramolecular polymerization: e.g. in both cases, monomers and polymer chains are present halfway through the polymerization [26]. Furthermore, cooperative effects have also been observed in conventional polycondensation reactions: Flory's “principle of equal reactivity” [44] is disobeyed if the electronic properties of the termini are changed in course of the chain growth, in particular when the reactivity of the polymer end groups becomes higher as for the monomer (i.e. the reaction of monomers becomes retarded). The polycondensation of phenyl 4‐(N‐alkylamino)benzoate in THF under basic conditions, initiated by phenyl 4‐nitrobenzoate, represents one example in this respect [116]: the reactivity of the remote phenyl ester moiety is strongly decreased as soon as the monomer is deprotonated; thus, individual monomers will not react with each other. The monomer anion will then only react with the initiator, thus generating an activated species which, in turn, contains phenyl ester moiety with increased reactivity (relative to the anionic monomer). Consequently, only the activated monomers will react, resulting in a polycondensation which shows a chain growth rather than a step‐growth behavior [117, 118].
The cooperative supramolecular polymerization can be induced by electronic (as shown in the previous example), hydrophobic, as well as structural effects (e.g. the formation of helical structures). In the latter case, the initial polymerization is thermodynamically rather unfavored when compared to the elongation phase, and thus, polymerization is preferred only when a critical length of the growing polymer chain is reached; then, due to conformational and/or structural changes, the growth of the polymer chain eventually becomes more favorable. As pointed out above, the nucleus is defined as the critical oligomer at which length the elongation of the polymer chain becomes more favored compared to dissociation. If cooperativity is arising (mainly) from structural changes, the nucleus can be regarded as the smallest possible species at which an unstructured, disordered assembly is transformed into an ordered one. A typical example for such a covalent polymerization reaction is the acid‐initiated polymerization of isocyanides, generating helical polymers. It has been proposed that the formation of an initial helical oligomer is required that, subsequently, serves as template for the attachment of further monomer units [119]. Remarkably, the anionic polymerization of triphenylmethyl methacrylate, using 9‐fluorenyllithium, as initiator, can be conducted asymmetrically when additional chiral ligands are present. Nakano et al. reported the synthesis of a one‐handed helical polymer [120]; the reactivity of each anionic oligomer was attributed to the DP and, thereby, to the specific chain conformation. The addition of further monomers to the chain occurred more readily when a stable helical conformation of the oligomer could be adopted (in this case, at a DP of 7–9).
Similarly, the Oya group reported on the cooperative chain growth of synthetic polypeptides via heterogeneous polymerization of amino acid anhydrides [121, 122]. Presumably due to the very short chains, preventing the formation of the anticipated R‐helical structure, antiparallel β‐sheet‐type oligopeptide species were initially observed; however, at a DP of c. 8, a conformational change into the R‐helix induced a more favorable chain growth via the addition of monomers to the active end of the helical structure.
Besides these representatives from the “covalent world,” various other examples for cooperative supramolecular polymerizations can be found. In particular, the structural cooperativity of H‐bonding interactions has been addressed by the Lehn [123], Meijer [124], and the Würthner groups [125] (see also Chapter 3). Other examples include the aggregation of cationic and anionic porphyrins (see Chapter 2) [126] as well as the stacking of oligo(p‐phenylene)s bearing dendritic ethylene glycol substituents [127], both driven by strong hydrophobic cooperativity effects. Moreover, metallophilic interactions have recently been identified to be the driving force in the supramolecular polymerization of oligo(phenylene ethynylene)‐based Pd(II)–pyridyl complexes [128] or linear Pt(II)–acetylide complexes (Chapter 4) [129]. In contrast, anti‐cooperative supramolecular polymerizations are rare: for instance, the polymerization of amphiphilic perylene derivatives in water [130] or cyclic peptides based on α‐ and ɛ‐amino acids [131].
1.4 Beyond Classical Supramolecular Polymerization
The three aforementioned mechanisms represent the classical examples for how supramolecular polymerization might proceed. Going beyond these, various strategies have been developed to circumvent some limitations arising from the traditional approaches [37]. In particular, the control over the molar mass and the dispersity remain as major challenges. The living supramolecular polymerization aims to adopt the key features from covalent living polymerizations, i.e. good kinetic control over the initiation and propagation steps, and to transfer these into self‐assembly processes [132]. Meijer and coworkers identified the NEP process (see Section 1.3.3) as the most appropriate mechanism for this purpose. It was shown that a finely tuned balance between attractive (i.e. a combination of various non‐covalent interactions) and repulsive forces (i.e. electrostatic interactions) was crucial to enable control over the supramolecular polymerization [133]. Based on an interplay between isodesmic and cooperative pathways, Ogi et al. realized that the supramolecular self‐assembly of a porphyrin monomer equipped with H‐bonding entities and hydrophobic alkyl chains into nanofibers of narrow length dispersities (Đ value of 1.10) [134]. In the same context, the “living crystallization,” i.e. the seeded growth of block copolymers into micrometer‐sized micelles, needs also to be mentioned [135]. In particular, the Manners and Winnick groups employed this strategy to assemble block copolymers in a highly controlled fashion.
A “supramonomer,” i.e. a monomer that was formed via non‐covalent interactions, can be polymerized using covalent or supramolecular polymerization techniques [37]. The covalent polymerization of preformed metal complexes into metallopolymers represents one of the most common applications in this respect (see also Chapter 4) [36, 136]. However, these polymerizations (as well as other examples, e.g. [137, 138]) do not offer sufficient control over the molar mass (distribution) since the reactions typically follow a step‐growth mechanism. Nonetheless, this approach provides an alternative to the direct supramolecular polymerizations and, thus, widens the options to prepare tailor‐made polymers that incorporate supramolecular units.
The self‐sorting of molecular components due to the selectivity of non‐covalent binding represents an established approach to assemble alternating supramolecular polymers, in particular when making use of orthogonal types of secondary interactions (see Chapter 11) [139]. Huang et al. demonstrated that supramolecular self‐sorting of two different cucurbit[n]uril derivatives (n = 7 and 8), in the presence of an appropriate homoditopic guest, can even proceed with reasonable control (see also Chapter 7) [140]. At an equimolar ratio of all three components, a maximum molar mass (Mw) of 9.7 × 104 g mol−1 was obtained; the dispersity of the polymer was c. 1.5 and thus remarkably low for such a supramolecular polymerization.
As one further example, the stimuli‐controlled supramolecular polymerization needs to be mentioned. Yang et al. proposed that supramolecular polymerizations, driven by external stimuli, might proceed with good control (assuming optimized reaction conditions) [37]. However, the broad range of stimuli‐responsive polymers reported so far basically rely on the ability to reversibly polymerize/depolymerize when, e.g. the pH‐switchability of ionic interactions (Chapter 2) or the redox‐switchability of host‐guest interactions (Chapter 7).
Not only supramolecular polymers are typically assembled under thermodynamic control, but