A Superhuman Sensing Triboelectric Nanogenerator with Boosted Power Density and Durability via a Bio-Inspired Janus Structure

Abstract

Triboelectric nanogenerators (TENG) not only enable sustainable self-powered sensing of devices, but also have superhuman noncontact/contact identification capabilities, which are propelling humanity toward the intelligent era. However, the inherently low dielectric constant of triboelectric materials as well as the mechanical mismatch between electrodes and dielectric materials severely limited their efficient and stable output performance. Taking inspiration from the asymmetric structure and function of human skin, a novel single-electrode TENG is developed, whose electrode and dielectric layer are integrated in a Janus architecture. By tuning the balance between gravity and the internal noncovalent interactions, gradient dispersion of carbon nanotubes in waterborne polyurethane networks can be feasibly achieved, which can boost the device performance by reinforcement in both the charge trapping capacity of the dielectric layer and the charge transfer of the electrode layer. As a proof-of-concept, triboelectric sensing and deep learning are integrated to realize the evolution from perception to identification under both noncontact (motion prediction) and contact (material identification) modes. The bionic design strategy of gradient Janus film can offer valuable insights into improving the output performance and durability of TENG. Additionally, the proximal prediction and tactile identification functions are also desirable attempts for future human-machine interfaces.