Electrical and Computer Engineering Faculty Publications
Document Type
Article
Publication Date
12-20-2025
Abstract
In this paper, we systematically demonstrate the design and analysis of a new type of ultra-broadband tunable metamaterial perfect absorber (MPA) comprising a top vanadium dioxide (VO2) based patterned resonating patch, a continuous metallic film at the bottom, and an intermediate dielectric substrate having a thickness of only 0.18 at the center working frequency. The simulation results reveal that the absorber achieves a bandwidth of 7.26 THz, ranging from 5.40 THz to 12.66 THz, with more than 90% absorptance and an average absorption of 98.21% under normal incidence of the incoming THz wave. Furthermore, absorptance exceeding 99% is achieved between 6.25 THz and 11.30 THz (5.05 THz bandwidth), demonstrating superior performance compared to existing broadband absorbers. The high absorption efficiency is attributed to the electric and magnetic dipole resonance, as illustrated through the electric field and vector surface current distribution at different frequencies. An equivalent RLC circuit model is developed using the least squares method, showing strong agreement with full-wave numerical simulations. However, designing metamaterial absorber requires extensive analysis of absorption spectra across a broad range of structural parameters – a computationally expensive process due to the complex interplay of impedance matching and electric field coupling. To overcome this challenge, we introduce a machine learning (ML)-based approach utilizing the Random Forest (RF) algorithm to predict absorption bandwidth and optimize structural parameters, significantly reducing computational time and spectral analyses. The RF model achieves considerably high accuracy, predicting an ultra-broadband absorption bandwidth of 7.26 THz with minimal error. We show that predicted and simulated results show excellent agreement, with negligible deviations. In addition, the terahertz absorber stably maintains more than 90% absorptance for both transverse electric (TE) and transverse magnetic (TM) waves up to 50° and due to its rotationally symmetric structure the proposed absorber is easy to fabricate and ensures complete polarization insensitivity. With its strong performance, the proposed MPA offers considerable potential for applications in terahertz modulation, switching, imaging, and biochemical sensing.
Recommended Citation
Ahmed, Md Rezwan, Oishi Jyoti, Pritu P. Sarkar, Mohammod Abdul Motin, Md Selim Habib, and Md Samiul Habib. "Accurate prediction of geometrical parameters of an ultra-broadband metamaterial absorber using machine learning." Scientific Reports 15, no. 1 (2025): 44263. https://doi.org/10.1038/s41598-025-27928-4
Publication Title
Scientific Reports
DOI
10.1038/s41598-025-27928-4
Comments
Student publication.
This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material.