Introduction
Surface Plasmon Resonance (SPR) biosensors have emerged as indispensable tools for real-time, label-free detection of biomolecules. However, conventional SPR sensors that employ noble metals are often hampered by substantial optical losses and constrained tunability in the near-infrared (NIR) spectrum. To address these challenges, we introduce an optimized SPR biosensor that integrates transparent conducting oxides (TCOs)—specifically, aluminum-doped zinc oxide (AZO) and indium tin oxide (ITO)—with graphene, thereby enhancing both plasmonic performance and biomolecular interactions.

Methods
The sensor architecture was developed using the Kretschmann configuration, incorporating a BK-7 prism along with sequential layers of AZO, ITO, graphene monolayers, and an additional dielectric layer to maximize sensitivity. Reflectance was modeled via the Transfer Matrix Method (TMM), while the dielectric characteristics of AZO and ITO were determined using the Drude–Lorentz oscillator model. Optimization of critical sensor parameters—including layer thicknesses and incident angles—was achieved through the application of machine learning techniques (Random Forest algorithms) and genetic algorithms (GAs).

Results
The proposed biosensor achieved a maximum sensitivity of 197.64°/RIU and a Figure of Merit (FOM) of 780.4 RIU⁻¹, thereby outperforming conventional SPR sensors based on gold and silver. The incorporation of graphene notably enhanced biomolecule adsorption, while the additional dielectric layer contributed to improved detection accuracy. Moreover, comparative analyses revealed that TCO-based SPR sensors exhibit markedly lower optical losses in the NIR range.

Conclusions
This study presents a novel and highly sensitive SPR biosensor that exploits the superior plasmonic properties of TCOs. By substituting traditional noble metals with AZO and ITO, the sensor achieves enhanced sensitivity, reduced optical losses, and superior detection accuracy. Furthermore, the integration of machine learning optimization techniques significantly refines sensor performance, paving the way for next-generation biosensors with broad applications in medical diagnostics, environmental monitoring, and biophotonics.