Eutectogel based on biomimetic heterophase liquid-locking design enables ultra-stable human-machine interaction interfaces

Conductive gel-based flexible wearable devices have emerged as transformative platforms for emerging applications in human-machine interaction. However, environmental humidity fluctuations can impair material performance through mechanisms such as water uptake-induced swelling/dehydration-induced embrittlement, as well as signal baseline drift. In this study, inspired by the strong hydration mechanism of natural moisturizing factors (NMF) and the sebum barrier in the skin, a biomimetic heterophase liquid-locking (BHL) design is proposed. This design precisely recapitulates the dual protective characteristics of skin. An internal hydrophilic phase, formed by zwitterionic polymers, achieves NMF-like high moisturization via hydrogen bonding/ion-dipole interactions. Meanwhile, a stable hydrophobic phase, mimicking the sebum barrier, is constructed on the surface through the covalent grafting of nanoscale hydrophobic polyhedral oligomeric silsesquioxane (POSS). The resulting gel exhibits ultra-stable mechanical and sensing performance over an ultra-wide humidity range (RH = 10%–90%), with less than 5% variation over 108 days. Furthermore, the eutectogel possesses softness, self-healing capability, high transparency, a wide temperature operating window, and ionic conductivity. By integrating gel-based electrode arrays and devices, human-machine collaboration and virtual reality (VR)-based human-machine interaction are realized. Moreover, an intelligent human-machine interaction system developed by further incorporating deep learning technology achieves a recognition rate of up to 99.68% for various gesture signals.