Optimization of Novel 2D Material Based SPR Biosensor Using Machine Learning.

Abstract

Biosensors are needed for today's health monitoring system for detecting different biomolecules. Graphene is a monolayer material that can be utilized to sense biomolecules and design biosensors. We have proposed a Graphene-Gold-Silver hybrid structure design based on Zinc Oxide which gives sensitive performance to detect hemoglobin biomolecules. The advanced biosensor designed based on this hybrid structure shows the highest sensitivity of 1000 nm/RIU which is far better concerning similar structure previously analyzed. The graphene-gold-silver hybrid structure is presented for its possible reflectance results and electric field results. The E-field results match well with the reflectance results given by the sensitive hybrid structure. The sensing biomolecules are presented above the structure where a combination of graphene-gold-silver hybrid structure improves the sensitivity to a great extent. The optimized parameters are obtained by applying variations in the physical parameters of the design. The machine learning algorithm employed for reflectance prediction shows a high prediction accuracy and can be utilized for simulation resource reduction. The proposed biosensor can be used in real-time hemoglobin monitoring.