.png){kind=icon}
.png){kind=icon}
Return
Machine Learning Assisted Analysis of Silver Nanorods Loaded Plasmonic Grating Based Sensor
.png){kind=icon}
PRE
AI
P
R
DOI：10.1007/s11468-026-03414-z.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Machine learning (ML) has become a potent tool for quick prediction of complex photonic or plasmonic systems. The approach drastically reduces the computational complexity of the conventional numerical simulation methods. In this work, a silver nanorod-loaded plasmonic grating is proposed for refractive index sensing. Plasmonic grating is made of metal-insulator-metal (MIM) waveguide to support high-level confinement. A wide bandgap based transmission behaviour is observed for the plasmonic grating due to destructive interference effect. A resonant peak is excited in the bandgap region once a defect cavity is introduced in the grating. When this defect cavity is loaded with silver nanorods, pronounced resonance tuning is achieved due to enhanced near-field coupling between nanorods. The coupling between rods is controlled by tuning the gap between silver rods. This effect is further investigated to sense the change in refractive index of the defect medium. The proposed device exhibits high sensitivity together with a high figure of merit (FOM) of 118.3, demonstrating its excellent sensing performance. All the numerical data is used to train different ML models, and the best performing model is identified based on mean square error and R2 score. The nonlinear relations between input features and output transmission is best captured by XGBoost model with MSE = 0.000649 and R2 score = 0.9734. Furthermore, the trained model is utilized for transmission prediction and ML-assisted design optimization for different geometrical parameters, thereby facilitating rapid sensitivity estimation. The obtained sensitivity through ML prediction is in good agreement with the sensitivity obtained through simulation tools. Therefore, the proposed hybrid strategy that utilizes amalgam of plasmonic-based simulation and ML approach is a viable solution for integrated photonic applications and optical sensing.
Keywords:
Machine learning
Plasmon grating
Metal-insulator–metal (MIM)
Sensitivity
Figure of merit
Journal
IF:
4.3
Papers:
4.3K
Citations:
7.5K
Organization
Cited Papers