Terahertz multi-resonator refractive index sensor with graphene and MXene integration for cancer biomarker analysis

In this study, we propose an innovative biosensor design that incorporates multiple resonators specifically engineered for detecting carcinoembryonic antigen (CEA). The biosensor integrates a unique combination of materials, including MXene, black phosphorus, and graphene, arranged in a hybrid configuration. The design consists of a centrally positioned circular MXene resonator encircled by a square ring of black phosphorus, complemented by four gold circular resonators. These components are assembled on a silicon dioxide substrate. A comprehensive performance evaluation was conducted across the terahertz spectrum (0.1–1.0 THz) using finite element method modeling in COMSOL Multiphysics 6.2. The biosensor demonstrated impressive metrics, including a maximum sensitivity of 811 GHz RIU⁻¹, a figure of merit of 14.479 RIU⁻¹, and a quality factor of 6.946. When tested with varying CEA concentrations ranging from 0 to 5 ng/mL, the device maintained stable and reliable operation. The transmission characteristics revealed systematic frequency variations from 0.389 THz to 0.382 THz. Additionally, a machine learning approach based on stacking ensemble regression was implemented to optimize sensor parameters. This computational strategy delivered outstanding predictive performance, achieving near-perfect accuracy across most operational variables. The biosensor's combination of high sensitivity, compact design, and reliable functionality positions it as a promising technology for early cancer screening and patient monitoring applications.