Simultaneous identification and detection of five antibiotics using a machine learning-enhanced glass capillary SERS...

Area of Science:

• Analytical Chemistry
• Environmental Science
• Materials Science

Background:

• Simultaneous detection of emerging contaminants like antibiotics in complex environmental samples is analytically challenging.
• Interference from sample complexity and overlapping spectral characteristics hinders practical applications.

Purpose of the Study:

• To develop a sensitive and selective platform for the simultaneous detection of multiple antibiotics in complex environmental matrices.
• To leverage surface-enhanced Raman scattering (SERS) within a microfluidic capillary format for enhanced analytical performance.

Main Methods:

• A glass capillary platform was functionalized with silver nanoparticles (AgNPs) stabilized by polyvinylpyrrolidone (PVP) to create a SERS-activated sensing surface.
• Microsampling was performed using the capillary, and SERS spectra were acquired for five different antibiotics.
• Principal Component Analysis (PCA) and Support Vector Machine (SVM) machine learning algorithms were employed for data analysis and classification.

Main Results:

• The SERS-activated capillary platform demonstrated a low limit of detection (LOD) for antibiotics as low as 7.62 × 10^-9 M.
• The developed method successfully distinguished and classified five different antibiotics, even in mixed samples and complex matrices.
• High classification accuracy of 99% was achieved using the SVM model with microliter-scale sample volumes.

Conclusions:

• The SERS-activated glass capillary sensor offers a sensitive and effective approach for the simultaneous detection of multiple antibiotics in complex environmental samples.
• The integration of PCA and ML algorithms enables robust data analysis and accurate identification of antibiotic mixtures.
• This technology presents a promising solution for environmental monitoring and contaminant analysis.