Internal Sensing with Exposed Core Fiber Plasmonic Sensor and Machine-Learning Approach for RI Prediction

Abstract

This study introduces, simulates, and evaluates an exposed core photonic crystal fiber (ECPCF)-based refractive index (RI) sensor where the surface plasmon resonance (SPR) phenomenon is incorporated through the noble metal gold. The sensor delivers exceptional performance, achieving a maximum wavelength sensitivity (W_S) of 23,000 nm/RIU and a figure of merit (FOM) of 287.50 RIU^-1. It covers an RI range of 1.33 to 1.41, making it suitable for identifying a wide variety of substances, including cancer cells, and biochemicals, demonstrating its versatility for optical sensing applications. Additionally, the incorporation of support vector regression (SVR) technique enhances accuracy and minimizes losses for RI prediction, offering promising advancements for applications such as lab-on-chip technologies. The findings highlight the sensor’s potential to revolutionize optical sensing with its remarkable sensitivity and extensive detection range.