A dual-network hydrogel sensor via stepwise hydrogen-bond modulation for sensing in underwater and sub-zero environments

Conductive hydrogels can convert external stimuli (such as strain and pressure) into detectable electrical signals, making them ideal candidates for next-generation sensors. Among them, dual-network (DN) hydrogel sensors display uniform stress dissipation, thus effectively enhancing the overall mechanical properties compared to single-network hydrogels. However, the construction of dual networks still faces challenges in reconciling the formation of physically and chemically crosslinked networks. Additionally, the presence of a large amount of unstable “free water” within hydrogels also limits their adaptability to multi-environment sensing. To address these issues, this study developed a stepwise hydrogen bond modulation strategy based on the Hofmeister effect for fabricating DN hydrogels. By introducing magnesium chloride (MgCl2) to weaken interchain hydrogen bonds, polyvinyl alcohol (PVA) hydrogels could achieve more efficient absorption of acrylamide (AAm) solution to form a highly interpenetrating DN. Meanwhile, MgCl2 and poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS) could synergistically enhance the ionic conductivity. Subsequently, the water/glycerol solution of sodium citrate (Na3Ct) could further strengthen the PVA cross-linking and modulate “free water”, thereby enhancing the mechanical strength and environmental tolerance. The study also provided an in-depth analysis of the underlying mechanism using Raman spectroscopy, offering new insights and theoretical support for related future research. The prepared hydrogel enhanced-PVA/PAAm/PEDOT:PSS/MgCl2 (PPPM-E) demonstrated capabilities in underwater Morse code communication and low-temperature gesture recognition sensing. Based on a convolutional neural network, the system achieved 99.56% accuracy in gesture recognition, highlighting its excellent multi-environment sensing performance.