柔性可穿戴设备在个人人体运动传感、智能疾病诊断和多功能电子皮肤等方面引起了广泛的兴趣。然而,已报道的柔性传感器大多表现出检测范围窄、灵敏度低、可降解性有限,加剧了大量电子废物对环境的污染,抗菌性能差,难以改善长期佩戴下细菌生长引起的皮肤不适和皮肤炎症。在此,受生物启发,人体皮肤具有高度敏感的触觉,具有棘状微结构,可放大表皮和真皮之间的传感灵敏度,采用可降解弹性体基材制备了一种可穿戴抗菌可降解电子产品,该基底具有以荷叶为模板的 MXene 涂层棘状微结构,并与叉指电极组装在一起。容易获得具有可调模量的可降解弹性体,以匹配人体皮肤的模量,并改善亲水性以实现快速降解。所获得的传感器具有超低检测限(0.2 Pa)、更高的灵敏度(高达 540.2 kPa −1 )、出色的循环稳定性(>23,000 次循环)、宽检测范围、强大的可降解性和优异的抗菌能力。在机器学习的推动下,从志愿者手指上的集成传感器收集的传感信号到相关的美国手语可以被有效识别,准确率高达99%,在无线人体运动传感和智能机器学习人机交互方面显示出巨大的潜力。相互作用。
"点击查看英文标题和摘要"
Flexible antibacterial degradable bioelastomer nanocomposites for ultrasensitive human–machine interaction sensing enabled by machine learning
Flexible wearables have attracted extensive interests for personal human motion sensing, intelligent disease diagnosis, and multifunctional electronic skins. However, the reported flexible sensors, mostly exhibited narrow detection range, low sensitivity, limited degradability to aggravate environmental pollution from vast electronic wastes, and poor antibacterial performance to hardly improve skin discomfort and skin inflammation from bacterial growth under long-term wearing. Herein, bioinspired from human skin featuring highly sensitive tactile sensation with spinous microstructures for amplifying sensing sensitivity between epidermis and dermis, a wearable antibacterial degradable electronics is prepared from degradable elastomeric substrate with MXene-coated spinous microstructures templated from lotus leaf assembled with the interdigitated electrode. The degradable elastomer is facilely obtained with tunable modulus to match the modulus of human skin with improved hydrophilicity for rapid degradation. The as-obtained sensor displays ultra-low detection limit (0.2 Pa), higher sensitivity (up to 540.2 kPa−1), outstanding cycling stability (>23,000 cycles), a wide detection range, robust degradability, and excellent antibacterial capability. Facilitated by machine learning, the collected sensing signals from the integrated sensors on volunteer's fingers to the related American Sign Language are effectively recognized with an accuracy up to 99%, showing excellent potential in wireless human movement sensing and smart machine learning-enabled human–machine interaction.