Interfacial mechanics are critical to the sensing performance of flexible strain sensors, influencing key metrics such as sensitivity, strain level, hysteresis, and long-term durability. Considering that the complex nature of multiscale interfaces within sensors and its sensing behavior arose from the overall response of interfaces rather than the individual interface, it presents a significantly experimental challenge to comprehensively elucidate their impact on sensor performance. Herein, we prepared microcrack-type MXene/carbon nanofiber/MXene-based polydimethylsiloxane strain sensors with varied interfacial interactions (i.e., strong, moderate, and weak) in the sensing layer by modifying the fabrication procedures. A combination of experiments and digital image correlation (DIC) was employed to investigate the role of these interfacial interactions on the strain field distribution. We found that the moderate interaction achieved an optimal balance between sensitivity and the strain range, making it the most suitable for sensing applications. To further enhance the sensitivity, microsphere arrays were incorporated into the PDMS substrate to induce highly localized strain fields, as evidenced by finite element analysis (FEA) and DIC. Benefiting from controllable crack initiation and opening, the resulting hierarchical strain sensor exhibited enhanced sensitivity without sacrificing the effective sensing region. These sensors were applied to monitor muscle activity in conjunction with machine learning techniques, enabling the accurate classification and prediction of muscle performance, thereby enhancing real-time feedback capabilities and providing personalized recommendations for optimizing rehabilitation and recovery.
Jarapanyacheep et al. (Thu,) studied this question.