Photoresponsive nanocomposite formed by self-assembly of an azobenzene-modified silane.

Mesoporous materials made by surfactant-directed selfassembly (SDSA) have been of great interest in the past decade because of their potential applications in catalysis, membrane separations, adsorption, and chemical sensing.[1–6] In SDSA, inorganic moieties formed by alkoxide hydrolysis interfacially coorganize with hydrophilic surfactant headgroups to form periodically ordered hybrid mesophases. After condensation and removal of surfactant molecules by heat or solvent extraction, a periodic mesoporous inorganic matrix is created. To expand the range of applications as well as create new properties, various organic functional groups have been covalently incorporated onto the pore surfaces of mesoporous materials, including phenyl, octyl, aminoalkyl, cyanoalkyl, thioloalkyl, epoxyl, and vinyl groups.[6–10] However, to date, these modifications have provided mainly “passive” functionality, such as controlled wetting properties, reduced dielectric constants, or enhanced adsorption of metal ions. By comparison, materials with “active” functionality would enable properties to be dynamically controlled by external stimuli, such as pH, temperature, or light. In an effort to make such active hybrid materials, Ogawa et al. intercalated azobenzene moieties within a clay (magadiite), which resulted in a hybrid layered composite in which restricted photoisomerization caused the d spacing to change reversibly by 0.6 / upon exposure to UV and visible light.[11] Several groups have reported[1213] or suggested the formation of photoactive hybrids through SDSA of azobenzene of pyridylethylene bridged silsesquioxanes. For a lamellar mesostructure, Brinker and co-workers. showed the d spacing to be optically controllable through photoisomerization of the azobenzene moiety before or after self-assembly. Alvaro et al.[13] doped 1,2-bis(4-pyridyl)ethylene in the wall of a hexagonal mesoporous silica and provided preliminary evidence of photoinduced changes in the surface area and pore volume. As shown in Figure 1, the incorporation of covalently bonded “switchable” ligands as pendant groups (as opposed to bridging groups[12–14]) within a porous (rather than a layered[1112]) framework is expected to result in nanocompo-