Machine-Learning-Guided Design of Ionic-Liquid-Modified Dual-Network Hydrogels for Flexible Electronics

Abstract

Conductive polymer hydrogels that integrate mechanical robustness, self-healing, adhesion, ionic conductivity, and long-term stability are highly desirable for flexible electronics. Herein, a machine learning (ML)-guided strategy was developed to design multifunctional polyacrylamide/chitosan dual-network hydrogels incorporating 1-ethyl-3-methylimidazolium tetrafluoroborate ([EMIM][BF4]) and sorbitol. An exploratory ML screening of 31 literature-derived additives prioritized imidazolium-based ionic liquids (ILs) as promising candidates within the investigated chemical space. The optimized hydrogel exhibited a tensile stress of 0.16 MPa, an elongation at break of 688.6%, a tensile-strength recovery of 92.7% after 30 min, an ionic conductivity of 3.2 S m–1, and a broad-range gauge factor (GF) of 4.277. In situ characterizations, component-removal experiments, density functional theory (DFT) calculations, and molecular dynamics (MD) simulations provided complementary evidence that hydrogen bonding, Al3+-mediated coordination, electrostatic interactions, and interfacial interactions collectively contribute to the polymer network and ion-transport behavior. The hydrogel enabled reliable human motion monitoring, handwriting recognition, human-machine interaction (HMI), and flexible supercapacitors with 85% capacitance retention after 10,000 cycles. This work provides a data-driven polymer-network design strategy for multifunctional conductive hydrogels toward flexible electronic applications.