Angle-robust hybrid 2D-material metasurface biosensor for COVID-19 detection with machine-learning optimization

This study presents a high-sensitivity metasurface-based biosensor for COVID-19 detection, incorporating advanced two-dimensional materials which includes graphene, borophene, MoS 2 , phosphorene, and germanene. The design consists of a graphene base layer supporting two phosphorene-coated rectangular resonators, a germanene-coated central circular resonator, and a MoS 2 -coated concentric ring structure. COMSOL Multiphysics simulations demonstrate an exceptionally high refractive-index sensitivity of 667 GHz RIU −1 within the 1.334–1.355 RIU range. Machine-learning-assisted optimization using polynomial regression enhances predictive reliability, achieving R 2 values between 87 % and 100 %. The proposed sensor exhibits a detection accuracy of 32.258 and a maximum figure of merit (FOM) of 21.505 RIU −1 , indicating strong potential for rapid and precise point-of-care COVID-19 diagnostics. Furthermore, the sensor maintains stable performance under varying incidence angles (0–80°) and tunable graphene chemical potentials (0.1–0.9 eV), confirming its robustness and practical adaptability.