Machine learning-assisted triple-emission Ln-MOFs sensor array for detection of multiple PFCs in aqueous environments

Area of Science:

• Environmental Chemistry
• Materials Science
• Analytical Chemistry

Background:

• Perfluorinated compounds (PFCs) are persistent environmental pollutants with potential carcinogenicity.
• Conventional single-emission probes struggle with sensitivity to environmental disturbances for PFC identification.

Purpose of the Study:

• Develop a robust, multi-channel fluorescence sensing platform for rapid and accurate PFC detection.
• Overcome limitations of traditional probes susceptible to environmental variables.

Main Methods:

• Constructed a three-channel fluorescence sensing platform using a europium/terbium bimetallic organic framework.
• Employed dual ligands (2,3,5,6-tetrafluoroterephthalic acid and 1,10-phenanthroline) for coordinated Eu³⁺/Tb³⁺ ions.
• Utilized hierarchical clustering, linear discriminant analysis, and machine learning for data analysis.

Main Results:

• Achieved triple characteristic emission (Eu³⁺, Tb³⁺, ligands) under single-wavelength excitation.
• Demonstrated a wide linear detection range (0.1–100 μM) for six PFCs with a low detection limit (42 nM).
• Accurately identified PFCs and mixtures, enabling concentration-independent qualitative and quantitative determination.

Conclusions:

• The developed platform provides rapid, high-precision monitoring of trace persistent pollutants.
• Multi-signal collaboration and intelligent analysis are significant for environmental risk management and pollution reduction.
• Blind sample validation confirmed the model's feasibility for practical applications.