Superior compressive and tensile bi-directional strain sensing capabilities achieved using liquid metal Hybrid-Hydrogels empowered by Machine learning algorithms

Abstract

Hydrogel-based flexible strain sensors have great potentials for applications in wearable electronics, human–machine interfaces, and soft robotics. Although great efforts have been made on enhancing tensile strain sensing capability of hydrogels, their sensing capabilities under compressive deformation have seldom been explored. Herein, we report a liquid metal-based hybrid hydrogel which achieves high stretchability (2300 %), superior bi-directional responses to both compressive strains (with a gauge factor of 33.43) and tensile strains (with a gauge factor of 9.59), fast response time (190 ms), and excellent durability (>1500 cycles). This conductive hybrid hydrogel with multi-interpenetrating networks is constructed by incorporating liquid metal, graphene oxide, poly-dopamine, and potassium chloride into a polymer double-network of polyacrylamide and poly(3,4-ethylenedioxythiophene): poly (styrene sulfonate). By integrating machine-learning algorithm with the hybrid hydrogel sensors, an intelligent dual-mode handwriting recognition system is developed for perceiving finger touch signals (in compressive-strain mode) and finger bending signals (in tensile-strain mode), with high accuracy (>93 %) and fast recognition time (<1s) when recognizing numbers, letters, and signature signals. Furthermore, this handwriting recognition system demonstrates advanced performance for human–machine interactions (e.g., playing music) and virtual reality interactions. This study offers a promising strategy to develop hydrogel electronics with bidirectional sensing capabilities for diversified practical applications.