Tracking natural organic matter (NOM) in a drinking water treatment plant using fluorescence excitation-emission matrices and PARAFAC

Natural organic matter (NOM) in water samples from a drinking water treatment train was characterized using fluorescence excitation emission matrices (F-EEMs) and parallel factor analysis (PARAFAC). A seven component PARAFAC model was developed and validated using 147 F-EEMs of water samples from two full-scale water treatment plants. It was found that the fluorescent components have spectral features similar to those previously extracted from F-EEMs of dissolved organic matter (DOM) from diverse aquatic environments. Five of these components are humic-like with a terrestrial, anthropogenic or marine origin, while two are protein-like with fluorescence spectra similar to those of tryptophan-like and tyrosine-like fluorophores. A correlation analysis was carried out for samples of one treatment plant between the maximum fluorescence intensities (F_max) of the seven PARAFAC components and NOM fractions (humics, building blocks, neutrals, biopolymers and low molecular weight acids) of the same sample obtained using liquid chromatography with organic carbon detection (LC-OCD). There were significant correlations (p < 0.01) between sample DOC concentration, UVA₂₅₄, and F_max for the seven PARAFAC components and DOC concentrations of the LC-OCD fractions. Three of the humic-like components showed slightly better predictions of DOC and humic fraction concentrations than UVA_254. Tryptophan-like and tyrosine-like components correlated positively with the biopolymer fraction. These results demonstrate that fluorescent components extracted from F-EEMs using PARAFAC could be related to previously defined NOM fractions and that they could provide an alternative tool for evaluating the removal of NOM fractions of interest during water treatment.