Machine learning–enabled direct ink writing of conductive polymer composites for enhanced performance in thermal management and current protection

Abstract

This study introduces a novel approach that leverages the synergy of machine learning (ML) and Direct Ink Writing (DIW) to optimize the manufacturing feasibility of conductive polymer composites (CPCs) films. The main research focus centers on precisely fine-tuning the printing parameters to strike the perfect balance between the high loading of dendritic copper fillers and their influences on processing and composite performance properties. This pioneering combination significantly enhances the precision of the final product without resorting to time-consuming procedures. ML algorithms contributed to identifying optimal printing variables such as speed, flow pressure, and filler concentration, which helped in identifying the optimal filler content and its performance capabilities. The resulting films exhibited thermo-resistive behavior with a noticeable resistivity increase by 6-7 orders of magnitude at elevated temperatures, specifically around 100 °C. Furthermore, the study highlights remarkable strain-sensing capabilities, simultaneously showcasing a substantial increase in composite modulus. These discoveries bear substantial significance for the development of exceptionally functional thermal interface materials suitable for use in sensors, current collectors, and energy storage devices. The method presented here offers a promising pathway for advancing the fabrication and performance optimization of conductive polymer composites, opening up diverse applications in emerging technologies.