Applying EEM-PARAFAC combined with moving-window 2DCOS and structural equation modeling to characterize binding properties of Cu (II) with DOM from different sources in an urbanized river

Abstract

Dissolved organic matter (DOM) in aquatic environment distinctly affects the behavior and fate of heavy metals via complexation, while the interfacial mechanisms and processes are still lacking in detail. Here, Cu (II) binding characteristics of DOM originated from hilly (NDOM), rural (RDOM) and urban (UDOM) regions in an urbanized river was explored by fluorescence excitation-emission matrix spectroscopy (EEM) combined with principal component coefficients, parallel factor analyses (PARAFAC), moving-window two-dimensional correlation spectroscopy (MW2DCOS) and structural equation modeling (SEM). Eight components were extracted from the titrants through EEM-PARAFAC, i.e., phenol-like substance (C1), tyrosine-like substance (C2), visible tryptophan-like substance (C3), ultraviolet tryptophan-like substance (C4), recent biological production (C5), wastewater-derived organic matter (C6), microbial humic-like substance (C7) and fulvic-like substance (C8). Interestingly, NDOM only contained C1, C3, C5 and C8, while nearly all components were found in RDOM (except for C2) and UDOM (except for C4). The f value of C1 (1.239) in NDOM was much higher than those in RDOM (0.134) and UDOM (0.115), so was of C8. It indicated that phenol-like and fulvic-like derived from autochthonous sources exhibited great binding ratios in the complexation with Cu (II). Moreover, C3 and C5 from UDOM exhibited higher f values (0.591 and 1.983) than those from NDOM and RDOM, suggesting that Cu (II) has a great binding capacity on protein-like from domestic and industrial wastewater. The MW2DCOS revealed that phenol-like and protein-like in NDOM and RDOM were essential for the binding of 160 μmol L⁻¹ Cu (II), whereas fulvic-like in NDOM and UDOM could react significantly with 10 μmol L⁻¹ Cu (II). Based on SEM, Cu (II) concentration had a negative direct effect on the fluorescence intensity of C7 or C8, whereas it showed an indirect positive effect on C7 or C8 through influencing C5, so was C6. It suggested that Cu (II) showed an indirect positive effect on the C8. This study might present a further comprehend of the environmental behaviors of Cu (II) in rivers.