Next-generation ultra-broadband and angularly stable MXene-based solar absorber with multilayer resonance engineering and AI-driven performance optimization

For sustained energy collection over a wide range of the spectrum, efficient solar thermal absorbers are crucial. In this research work, a multilayer MXene-based solar absorber is designed and systematically investigated to achieve ultra-broadband and high-efficiency absorption. The proposed structure consists of MXene and SiO 2 , with MXene used in a patterned resonator layer, a dielectric SiO 2 spacer, and a continuous MXene ground layer that suppresses transmission. Finite Element Method (FEM) numerical simulation results demonstrate near-unity absorption across the 200–2500 nm wavelength range, with a peak absorptance approaching 100% and an average absorption of approximately 96%. Parametric optimization analysis reveals that variations in structural length, resonator thickness, spacer thickness, and ground layer thickness have minimal impact on overall absorption, confirming strong design robustness and fabrication tolerance. Angular analysis shows stable absorption performance up to large incident angles, ensuring effective solar harvesting throughout the day without tracking systems. Electric field distributions at representative wavelengths illustrate multiple absorption mechanisms, including edge localization, interlayer coupling, and cavity resonance within the MXene layer. Furthermore, a machine learning (ML)–assisted optimization framework is employed to accurately predict absorption behaviour, achieving high R 2 values and low prediction errors, thereby significantly reducing computational effort. The combined numerical and data-driven results confirm that the proposed MXene absorber is a promising candidate for high-performance solar thermal and renewable energy applications.