Enhanced Terahertz Surface Plasmon Resonance Biosensor with Graphene-Metal Hybrid Architecture for Non-Invasive Glucose Detection

This study presents a biosensor architecture that integrates graphene with silver and bismuth in a multi-resonator architecture optimized for glucose detection. This work uniquely integrates graphene with silver and bismuth in a multi-resonator SPR configuration, demonstrating unprecedented sensitivity and linearity for non-invasive glucose detection. Unlike previous single-material or binary-material approaches, our design introduces a three-metal hybrid architecture (Ag/Bi/Au) combined with graphene in a geometrically optimized multi-resonator array specifically engineered for terahertz glucose sensing. Computational electromagnetic simulations using finite element methods reveal that the sensor achieves a sensitivity of 2000 GHz RIU −1 within the refractive index range of 1.335–1.347, which is particularly suitable for glucose monitoring applications. Key parameters including graphene chemical potential, angle of incidence, and resonator dimensions were systematically varied to optimize performance, with the sensor exhibiting a quality factor of approximately 10.8 and a figure of merit reaching 27.778 RIU −1 . The relationship between resonance frequency and refractive index demonstrates strong linearity (R 2 = 85%). Additionally, a stacking ensemble regression model was developed to predict absorption characteristics, achieving prediction accuracy up to 100% for certain parameter ranges. This research advances the development of high-performance terahertz biosensors for non-invasive glucose monitoring with potential applications in diabetes management.