Scholarlink

<p>This investigation presents a surface plasmon resonance biosensor architecture integrating graphene and gold metasurfaces, optimized through advanced machine learning methodologies. The biosensor design features engineered configuration of T-shaped, semi-circular ring, and circular resonators fabricated on a silicon dioxide (SiO<sub>2</sub>) substrate. Numerical simulations conducted via COMSOL Multiphysics software validate the sensor’s exceptional detection capabilities for both gaseous and aqueous analytes. The device demonstrates remarkable sensitivity metrics of 300 GHzRIU<sup>−1</sup> for gaseous substances and 200 GHzRIU<sup>−1</sup> for aqueous solutions. Systematic parametric optimization significantly enhances the biosensor’s performance characteristics. To further advance the predictive capabilities of the system, a one-dimensional Convolutional Neural Network (CNN) regression model is implemented, and the results demonstrate an <i>R</i><sup>2</sup> coefficient of 100%. The biosensor exhibits superior detection metrics, with figure of merit values reaching 4.225 RIU<sup>-1</sup> and quality factors of 11.981, establishing its viability for environmental monitoring applications and biomedical diagnostics.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Design and Optimization of Graphene-Gold Metasurface THz Biosensor Using Au-SiO2 Material with Machine Learning for Multi-Analyte Detection

  • 期刊论文

  • 发表日期：

    2025年5月5日

查看全文
  • Jacob Wekalao,
  • Habib Kraiem,
  • Sana Ben Khalifa,
  • Saleh Chebaane,
  • Ammar Armghan,
  • Shobhit K. Patel

摘要

This investigation presents a surface plasmon resonance biosensor architecture integrating graphene and gold metasurfaces, optimized through advanced machine learning methodologies. The biosensor design features engineered configuration of T-shaped, semi-circular ring, and circular resonators fabricated on a silicon dioxide (SiO2) substrate. Numerical simulations conducted via COMSOL Multiphysics software validate the sensor’s exceptional detection capabilities for both gaseous and aqueous analytes. The device demonstrates remarkable sensitivity metrics of 300 GHzRIU−1 for gaseous substances and 200 GHzRIU−1 for aqueous solutions. Systematic parametric optimization significantly enhances the biosensor’s performance characteristics. To further advance the predictive capabilities of the system, a one-dimensional Convolutional Neural Network (CNN) regression model is implemented, and the results demonstrate an R2 coefficient of 100%. The biosensor exhibits superior detection metrics, with figure of merit values reaching 4.225 RIU-1 and quality factors of 11.981, establishing its viability for environmental monitoring applications and biomedical diagnostics.