Automated machine learning assisted fluorescent sensor array based on silver nanoclusters for detection of multiple...

Area of Science:

• Analytical Chemistry
• Materials Science
• Environmental Science

Background:

• Heavy metal determination is challenging due to overlapping fluorescence responses.
• Developing selective and sensitive detection methods for multiple metal ions is crucial for environmental monitoring.

Purpose of the Study:

• To construct a multichannel fluorescent sensor array for simultaneous detection of multiple heavy metal ions.
• To apply an automated machine learning framework for analyzing complex fluorescence data and improving detection accuracy.

Main Methods:

• Fabrication of a sensor array using three types of silver nanoclusters (AgNCs) with distinct surface ligands.
• Utilizing an automated machine learning framework (AutoGluon) for spectral fingerprint analysis, including feature selection, classification, and regression.
• Testing the sensor array with seven common metal ions at various concentrations and in real environmental samples (lake water, soil extracts).

Main Results:

• The automated machine learning workflow achieved 100% classification accuracy for single-ion samples.
• The system enabled accurate ion-specific concentration prediction from multichannel spectral fingerprints.
• The sensor array successfully discriminated mixed metal ion samples and performed reliably in spiked environmental matrices, demonstrating low-level quantitative detection capabilities.

Conclusions:

• Combining cross-reactive AgNCs with automated machine learning offers a practical and effective framework for heavy metal detection.
• This approach enables rapid qualitative identification and quantitative prediction of multiple heavy metal ions in diverse samples.
• The developed method shows significant potential for real-time environmental monitoring and analysis.