Revolutionizing Electromagnetic Materials: Machine Learning Enabled Optimization of Polymer Nanocomposites for Enhanced Performance

Abstract

The demand for polymer nanocomposites (PNCs) has been steadily increasing in various electronic and electromagnetic applications. PNCs have played a crucial role in the development of radio frequency devicesand stealth technology resulting in a significant importance on microwave absorption materials. The aim of this work is to employ Gaussian process regression (GPR) as a powerful approach for multidimensional optimization of key parameters, such as filler content and thickness of PNC, in order to realize its reflection loss (RL) characteristics, viz., minmum RL (RLmin) value and RL ≤ −10 dB bandwidth (corresponds to 90% absorption). As a proof of principle, the solution‐processed polydimethylsiloxane (PDMS)–Fe2O3 nanocomposite is explored. Using experimental data from four different filler contents, the continuous electromagnetic response using GPR is predicted. By integrating an optimization algorithm with the GPR‐predicted electromagnetic responses, exceptional results in the form of an optimal RLmin value of −67 dB for a 1.4 mm thicker PDMS nanocomposite containing 33 vol% Fe2O3 nanoparticles have been achieved. This performance has not been reported previously, making it a significant contribution to the field. The experimental results corroborate the predicted data, providing evidence for the efficacy of this novel approach in designing robust PNCs for enhancing RL performance.