Synergistic optimization strategy for high-performance multilayer SPR biosensor

Surface plasmon resonance (SPR) biosensors are widely employed for label-free, real-time biomolecular detection, yet their performance is markedly limited for low-molecular-weight analytes such as urinary glucose. Meanwhile, conventional structural designs mostly rely on empirical guidance or single-parameter scanning, which constrains the optimization of overall device performance. To solve these challenges, we propose a high-sensitivity SPR biosensor with a ZnO/Au/Ag/Si/WSe₂ multilayer configuration for glucose detection in diabetes diagnosis. The structure leverages the complementary roles of functional layers to simultaneously enhance refractive index sensitivity and biomolecule adsorption efficiency. We also established a genetic algorithm-based multi-objective optimization method to globally optimize layer thicknesses, with sensitivity (S), Figure of merit (FOM) and resonance peak depth as core evaluation metrics. Simulation results show that the optimized sensor reaches a sensitivity of 362.96 °/RIU, an FOM of 60.00 RIU^-1 and a Depth figure of merit (DFOM) of 33.28 RIU^-1, which are 2.05 times, 2.73 times and 1.56 times higher than those of traditional Cr/Au-based SPR sensors, respectively. For urinary glucose samples with concentrations of 0-10 g/dL (refractive index range 1.335-1.347), the optimized structure achieves a maximum sensitivity of 284.85 °/RIU, an FOM of 60.10 RIU^-1 and a DFOM of 42.15 RIU^-1. These results confirm that the proposed biosensor has great application potential in noninvasive diabetes diagnosis, early cancer detection, food safety testing and environmental sensing.