Ultrawideband Multilayer Metal–MXene–Dielectric Solar Absorber: Numerical Design and Machine Learning–Driven Performance Prediction

Authors:
Open Access
Publisher website
Google Scholar
Abstract
This work presents a compact multilayer solar absorber based on a spiral multi‐ring resonator integrated with MXene for broadband energy harvesting. The proposed metal–dielectric–MXene–metal configuration consists of a patterned Fe resonator placed over a SiO 2 spacer and MXene layer, backed by a metallic ground plane to suppress transmission. The structure achieves near‐perfect absorption, reaching 99.9% peak and maintaining absorption above 0.85 across 0.3 to 3.3 µm. Parametric analysis reveals that MXene enhances absorption through strong electromagnetic coupling and intrinsic losses, while SiO 2 thickness governs spectral tuning. The design maintains stable absorption up to 60° incidence for both TE and TM modes, with superior TM performance. Broadband behavior arises from impedance matching, plasmonic resonances, and dielectric interference. Strong field confinement in the spiral resonator enhances absorption around 1.8–2.0 µm. Machine learning models predict absorption effectively, with AdaBoost achieving the highest accuracy (R 2 ≈ 0.943), demonstrating scalability and application potential.

This publication has 29 references indexed in Scilit: