A Trimodal 2D Metasurface Biosensor with Bayesian Regression for Ultra-Sensitive Cancer Biomarker Detection

In this study, we present the design, simulation, and characterization of a biosensor architecture leveraging a trimodal integration of MXene, black phosphorus, and graphene for the enhanced detection of neoplastic biomarkers. The sensor configuration consists of MXene-functionalized rectangular resonators coupled with black phosphorus-augmented circular ring structures, all integrated atop a graphene-modified substrate. The entire device was simulated on a silicon dioxide platform using standard photolithographic techniques. Electromagnetic performance, evaluated through finite element method (FEM) simulations (COMSOL Multiphysics), demonstrates outstanding sensing capabilities, with the sensor achieving an exceptional sensitivity of 2000 GHzRIU⁻¹ within the clinically relevant refractive index range of 1.36–1.401.A series of parametric studies were conducted to investigate the effects of key factors, including graphene chemical potential (GCP), incident electromagnetic wave angle, rectangular resonator dimensions, and circular ring radius on the transmission spectra. The sensor exhibits a strong linear relationship between resonance frequency shift and refractive index variation (R² = 98.918%). Moreover, Bayesian regression modelling applied to variations in GCP and incident angle yielded high predictive accuracy, with coefficients of determination of approximately 90% and 91%, respectively.