Activation-Free and Strain-Resilient Island-Bridge Liquid Metal Inks for Machine-Learning-Enabled Epidermal...

Liquid metal particle-based inks offer a scalable route toward stretchable electronics, yet their practical implementation is hindered by activation-dependent conductivity and poor electrical stability under deformation. Herein, we report an intrinsically conductive and strain-resilient liquid metal ink with a rationally engineered island-bridge micro-architecture, in which Cu-decorated carbon nanotubes (CuCNTs) bridge adjacent liquid metal particles (LMPs) through initial hydrogen-bonding/coordination-assisted attachment to the native gallium-oxide shell, followed by local Cu-Ga alloying at the CuCNTs/LMPs interface. Enabled by solvent-mediated dispersion and interfacial hybridization, the strain-resilient composite ink supports high-resolution (100 µm) screen printing and deformable circuit integration. The dynamically reconfigurable island-bridge network retains a sheet resistance of 1.3 × 10^-1 Ω sq^-1 at 300% strain, together with stable resistance during prolonged cyclic stretching and bending. When printed into a compact 12-channel array on breathable fibrous substrates, the resulting electrode array exhibits low skin-electrode impedance and superior high-density surface electromyography (HD-sEMG) performance compared with commercial gel electrodes. Combined with machine-learning-assisted analysis, the electromyography arrays enable training-state recognition, individualized load assessment, and continuous monitoring of fatigue-associated exercise states during dynamic exercise. This work provides a scalable strategy for intrinsically conductive liquid metal electronics and advances their application in intelligent wearable biointerfaces and soft electronic systems.