Molecularly Engineered Self-Adhesive Conductive Hydrogel Enabled by Multiple Non-Covalent Interactions for Facial Expression Recognition and Intelligent Grasping

Abstract

Flexible electronics and wearable devices require hydrogels that integrate high strength, stretchability, adhesion, and stable conductivity. However, conventional polyacrylamide (PAM) hydrogels suffer from poor mechanics, low conductivity, and weak adhesion, limiting their use in sensitive sensing. Herein, guided by molecular engineering, we design a ternary composite conductive hydrogel (PAM-I-L) by incorporating inositol and the ionic liquid [BSO3HMIm][OTf] into a PAM network, creating a multifunctional noncovalent network that synergistically boosts mechanical, adhesive, and electrical properties. Inositol boosts tensile strength from 75.6 ± 8.5 to 212.3 ± 16.2 kPa and fracture strain from 636 ± 16% to 1660 ± 111%, while [BSO3HMIm][OTf] enhances conductivity from 33.6 mS m–1 to 1090 mS m–1 and glass adhesion from 56.4 ± 2.2 to 130.2 ± 9.3 kPa. The hydrogel sensor exhibits a gauge factor of 2.2–8.2 over 0–600% strain, demonstrating excellent strain responsiveness. The sensor enables high-fidelity recognition of facial microexpressions and joint movements. When integrated as a multichannel sensor array on the inner side of each finger joint of a robotic hand, the system dynamically monitors force distribution during the grasping of objects with varying weights, achieving high signal consistency and repeatability across individual fingers. This validates its potential for applications in flexible robotics, intelligent grasping, and human-machine interaction.