Search

To rapidly monitor the surface water quality in terms of organic pollution of an industrial river undergoing restoration, optical methods (UV-visible spectrometry and fluorescence) were applied in parallel to classical physical-chemical analyses. UV-visible spectra were analyzed using the maximum of the second derivative at 225 nm (related to nitrates), specific absorbance at 254 nm (SUVA 254 ), and the spectral slope between 275 and 295 nm (S 275-295 ) (related to the aromaticity and molecular weight of dissolved organic carbon). The synchronous fluorescence spectra (wavelength difference = 50 nm) exhibited a high variability in the composition of dissolved organic material between the upstream and downstream sections and also versus time. The principal components analysis of the entire set of synchronous fluorescence spectra helped to define three river sections with different pollution characteristics. Spectral decomposition was applied to the two most upstream sections: five fluorophores, classical in rivers impacted by domestic sewage and related to protein-like (λ ex  = 280 nm) and humic-like fluorescence (M-type with λ ex  ≈ 305-310 nm and C-type with λ ex  ≥ 335 nm), were identified. The irregular shape of the synchronous fluorescence spectra in the most downstream section is likely due to organic pollutants of industrial origin; however, their variability and the complexity of the spectra did not allow the further elucidation of their nature.

In the last 20 years, increasing dissolved organic carbon (DOC) concentrations have been observed in several rivers and lakes in Europe. This increase has reduced the quality of the aquatic environment. In this study, UV-vis spectroscopy and synchronous fluorescence spectroscopy with a difference of 50 nm between the excitation and emission (SF50) were used to characterize the DOC in a rural river (Madon River). The specific absorbance index at 254 nm (SUVA254) which is related to the aromaticity of DOC was extracted from UV-vis spectra, whose maximum of the second derivative (occurring near 225 nm) is related to nitrates. SF50 spectra which are characterized by well-defined peaks indicated large spatial and temporal variations. Two methods were used to analyze and compare these spectra. The first method was based on the decomposition of the SF50 spectra into four Gauss functions: B1 (related to tryptophan-like fluorescence), B2 and B3 (related to humic substances), and B4 (related to chlorophyll-like substances). The second method was principal components analysis (PCA), which results yielded three principal components that accounted for 95% of the variance. Although PCA enables the consideration of the spectra without making assumptions regarding the number of fluorophores, the results from the decomposition in Gauss function were easier to interpret.

The photocatalytic degradation of a series of six acid dyes (Direct Red 80, Direct Red 81, Direct Red 23, Direct Violet 51, Direct Yellow 27, and Direct Yellow 50) has been tested compared in terms of color removal, mineralization, and toxicity (Lactuca sativa L. test) after photocatalysis on immobilized titanium dioxide. The dyes were examined at their natural pH and after hydrolysis at pH 12. Results show that hydrolysis decreases strongly the efficiency of color removal, that full mineralization takes much longer reaction time than color removal, and that toxicity is only very partially reduced. Some structural parameters, related to the structure and the topology of the dye molecules, could be correlated with the apparent color removal rates at natural pH.