▎ 摘　　要

The filler network in smart nanocomposites is capable of dictating the strain actuation and electric generation behaviour in shape-memory and dielectric-elastomer systems, respectively. Noteworthy, the dissipation parameters, such as loss modulus and dielectric loss, are the critical criteria for designing an ideal smart polymeric composite. For manipulating dissipation parameters, we developed three types of nanofiller, including (I) simple graphene-oxide (GO), (II) reduced graphene-oxide (rGO), and (III) noncovalent-factionalized graphene with (polyamine-anchored)-perylene-bisimide (XGO). After fabrication of poly-urethane (PU) nanocomposites at various filler loading, the robust correlation between microstructure, electrical and mechanical, including static and dynamic (linear and nonlinear viscoelasticity), properties of nanocomposites has been deduced. These smart polymeric nanocomposites demonstrate the capability of shape-memory actuating as well as harvesting the electric energy from mechanical work. First and foremost, the harvested-energy-density of PU was meaningfully improved when blended with XGO. A composite-film containing 5 wt% XGO with the harvested energy density of 2.97 mJ/cm(3) was achieved, which is about 6.7 times superior to that of pristine PU. Furthermore, the shape-memory recovery ratio of functionalized 2 wt%-nanocomposite significantly improved from 86.2% for pure PU to 93.4% for the XGO sample. The observations in this work strongly suggest compositing is an auspicious-way to provide proper shape-memory actuator and better dielectric-elastomer candidates for forthcoming practical generators.