Development of hybrid composites with zero-temperature coefficient of resistance: Mechanisms and predictive modeling using symbolic regression

초록

Carbon-based polymer composites, such as carbon black (CB) and carbon nanotubes (CNTs), have been extensively studied for applications in Joule heating-based heaters, strain sensors, and wearable devices. However, the relatively high coefficient of thermal expansion (CTE) of polymers can cause carbon fillers to displace during heating, resulting in resistance variations that reduce device accuracy and increase the risk of fire due to overcurrent. Specifically, polymer composites with zero-dimensional fillers exhibit a positive temperature coefficient of resistance (PTC), whereas those with one-dimensional fillers show a negative temperature coefficient of resistance (NTC). In this study, we developed hybrid composites with a zero-temperature coefficient of resistance (ZTCR) to maintain a stable resistance across a wide temperature range. By combining CNTs with NTC properties and CB with PTC properties, we achieved a compensatory effect that stabilized the resistance of the composites. We thoroughly investigated the mechanisms behind this effect and proposed a predictive model for ZTCR using symbolic regression in machine learning. This innovative approach ensures consistent electrical properties under varying temperature conditions, making these composites highly suitable for applications requiring precise resistance control.

키워드