Chromium Speciation Monitoring Platform for Drinking Water: Machine Learning-Assisted Dual-Emission Fluorescence Sensor Array

Different chromium (Cr) speciation in drinking water shows distinct risk levels to humans, failing to reflect real environmental impacts only by total Cr analysis. Integrated with machine learning, a novel fluorescence sensor array was developed for rapid identification and quantitative detection of Cr speciation without sample pretreatment other than filtration. This system prepared three-component fluorescence hybrid materials (MSN@Zr@Au and MSN@Zr@AgAu) with dual emission wavelengths. The sensing unit with a dual-mode algorithm was specific for Cr speciation and accurately identified chromium speciation among 11 coexisting cations. The algorithm of linear discriminant analysis (LDA) assisting hierarchical cluster analysis (HCA) provided higher selectivity for Cr speciation for real samples. Finally, this method showed good analytical performance ranging from 1 to 60 μM, exhibiting a low detection limit of 1.29 μM. This strategy shows excellent practicability for Cr speciation analysis in drinking and tap water, developing a practical monitoring platform for real water.