Liquid Metal and Carbon Nanofiber-Based Strain Sensor for Monitoring Gesture, Voice, and Physiological Signals

Abstract

Fiber strain sensors are emerging as a focus in wearable monitors due to their excellent flexibility and weavability. However, a compromise of the detection range with sensitivity and hysteresis still exists in current fiber strain sensors. To overcome this deficiency, we propose a fiber strain sensor by using liquid metal (LM) and carbon nanofiber (CNF) composites (LM/CNF) as elastic and sensitive conductive materials and dip-coating it on a polyurethane elastic thread (PUT) by an ultrasonic-assisted method. It is demonstrated by morphological examination that there is an "island-bridge" like the network in the LM/CNF suspension and the coated LM/CNF on PUT by the ultrasonicassisted dip-coating technique. The good sensitivities in a wide strain range are testified (i.e., gauge factors (GFs) of 14.8 ± 0.5, 35.7 ± 2.3, and 73.7 ± 3.9 in the strain ranges of 0−110, 110−150, and 150−220%, respectively) and proven to be contributed by the entangled LM nanoparticles by CNF microrods. Meanwhile, the short response times (304 ± 26 ms for loading 20% strain and 315 ± 28 ms for unloading) and low hysteresis are also manifested. All of these good features endow the LM/CNF-based fiber strain sensor with the capability to simultaneously monitor human health state and body movements; this great potentiality is also demonstrated by the on-body detections of respiration, pulse, voice, and joint bending. Finally, its conceptual application as a smart glove for gesture recognition is also carried out. In general, this work manifests that the LM/CNF-based fiber strain sensors may be of practical value in the field of health and motion detection.