No abstract

Traditional aqueous self-assembly of tubular structures (as well as other aggregates) usually relies on the hydrophobic effect, a relatively weak and nondirectional interaction. The resultant aggregates are inherently soft, fluid, and less-ordered. Alternatively, we report a novel kind of nonamphiphilic selfassembly of microtubes in aqueous solutions of cyclodextrin/ionic surfactant (CD/IS) complexes. This self-assembly is driven exclusively by H-bonds, relatively strong, directional interactions. The CD/IS microtubes feature an unbundling nature, ultralong persistence lengths, highly monodispersed diameters, and remarkable rigidity. Every single CD/IS microtube is constituted by a set of coaxial, equally spaced, hollow cylinders, resembling the annular rings of trees (thus termed as ''annular ring'' microtubes). Furthermore, bearing in mind the fundamental difference between the amphiphilic counterpart in driving forces, this H-bond-driven hydrophilic self-assembly is envisioned to complement its counterpart and expand the field of molecular self-assembly.

Proteins can readily assemble into rigid, crystalline and functional structures such as viral capsids and bacterial compartments. Despite ongoing advances, it is still a fundamental challenge to design and synthesize protein-mimetic molecules to form crystalline structures. Here we report the lattice self-assembly of cyclodextrin complexes into a variety of capsid-like structures such as lamellae, helical tubes and hollow rhombic dodecahedra. The dodecahedral morphology has not hitherto been observed in self-assembly systems. The tubes can spontaneously encapsulate colloidal particles and liposomes. The dodecahedra and tubes are respectively comparable to and much larger than the largest known virus. In particular, the resemblance to protein assemblies is not limited to morphology but extends to structural rigidity and crystallinity—a well-defined, 2D rhombic lattice of molecular arrangement is strikingly universal for all the observed structures. We propose a simple design rule for the current lattice self-assembly, potentially opening doors for new protein-mimetic materials.