Innovative Technologies for the Online Monitoring of Chloroform in Water by MicroNIR/Chemometrics Systems

This study evaluates a miniaturized Near-infrared spectroscopy MicroNIR/Chemometrics platform as a rapid tool for monitoring chloroform, a representative disinfection by-product from the trihalomethanes class. Two MicroNIR devices were tested in three configurations: reflectance measurements in glass vials and transmittance measurements in static and flow modes. Experiments were conducted with spiked distilled water over the 1–200 µg/L range relevant to regulations. Principal Component Analysis provided clear separation between blanks and chloroform-containing samples in all configurations. Partial Least Squares Discriminant Analysis achieved 100% specificity and non-error rate, with a root mean square error in prediction (RMSEP) of approximately 0.06–0.08 for static and flow measurements and 0.09–0.10 for vial mode. In the 1–20 µg/L range, Partial Least Squares Regression demonstrated comparable performance for static and flow measurements (R2 = 0.91; RMSEP = 2.3 µg/L), while flow provided the lowest estimated limits of detection and quantification (2.6 µg/L and 7.9 µg/L, respectively). In the 20–200 µg/L range, flow showed the highest quantitative performance. The results support the feasibility of MicroNIR/Chemometrics as a rapid screening tool for chloroform in water, complementary to reference methods. Future work will focus on improving inter-campaign transferability, extending the approach to all regulated trihalomethanes, and validating it in real matrices.