Reactive silver ink-derived SERS-active nanofiber platform enabling machine learning-assisted sensitive amino acid...

Area of Science:

• Analytical Chemistry
• Materials Science
• Spectroscopy

Background:

• Surface-Enhanced Raman Spectroscopy (SERS) offers sensitive analyte detection but faces challenges in substrate fabrication and spectral interpretation.
• Practical and scalable SERS substrates are crucial for widespread application in chemical analysis.

Purpose of the Study:

• To develop a scalable and practical machine learning-assisted SERS platform for sensitive analyte detection and classification.
• To create a novel SERS substrate using silver nanoparticle-decorated nanofibers (AgNFs).

Main Methods:

• Fabrication of SERS substrates by integrating silver ink into a nanofiber matrix (PVP/PEO) for guided nanostructure formation.
• Utilized Surface-Enhanced Raman Spectroscopy (SERS) for detecting amino acids at low concentrations (0.3 μM).
• Applied machine learning (ML) algorithms, including linear support vector machine (SVM), for spectral data analysis and amino acid classification.

Main Results:

• Achieved a high analytical enhancement factor (AEF) of 2.9 × 10⁷ with the AgNF SERS substrate.
• Demonstrated sensitive detection of five amino acids with a strong correlation between signal intensity and concentration (R² > 0.9).
• Attained approximately 98% classification accuracy for amino acids using the linear SVM algorithm.

Conclusions:

• The developed ML-enhanced SERS platform is effective for detecting and classifying various analytes with simplicity, scalability, and reproducibility.
• The AgNF-based SERS substrate offers ultra-sensitivity and practical applicability for chemical sensing.
• This integrated approach overcomes limitations in SERS substrate fabrication and spectral analysis, paving the way for advanced analytical tools.