Machine learning-enhanced surface plasmon resonance based photonic crystal fiber sensor.

Area of Science:

• Optoelectronics
• Nanotechnology
• Sensor Technology

Background:

• Surface Plasmon Resonance (SPR) is a powerful sensing technique.
• Photonic Crystal Fibers (PCFs) offer unique light-confining properties.
• Integrating SPR with PCFs enhances sensing capabilities.

Purpose of the Study:

• To introduce a novel photonic crystal fiber sensor with enhanced SPR capabilities using gold nanowires.
• To employ machine learning (Artificial Neural Networks - ANN) for predicting sensor performance metrics.
• To evaluate the sensor's sensitivity and the ANN models' prediction accuracy.

Main Methods:

• Fabrication of a photonic crystal fiber sensor with four gold nanowires.
• Utilizing Surface Plasmon Resonance (SPR) principles for sensing.
• Applying Artificial Neural Networks (ANN) to predict confinement loss and sensitivity based on sensor parameters.

Main Results:

• ANN models achieved high accuracy in predicting sensor outputs with low mean squared errors (0.084, 0.002, 0.003).
• The sensor exhibited significant wavelength sensitivities ranging from 2000-18000 nm/RIU for refractive indices between 1.31-1.4.
• A maximum amplitude sensitivity of 889.89 RIU^-1 was recorded.

Conclusions:

• The novel PCF-SPR sensor integrated with gold nanowires demonstrates excellent sensing performance.
• Machine learning (ANN) provides a reliable and efficient method for predicting optical sensor outputs.
• This approach highlights the potential of combining advanced materials, optical sensing, and AI for next-generation sensor design.