Study Of Complexation Of Cupric Ion (Ⅱ) By Dissolved Organic Matter (DOM) Based On Molecular Spectroscopy And Chemometrics

A fluorescence quenching model using Cu2+ ion selective electrode (Cu-ISE) is developed. By fitting the fluorescence intensity vs. [Cu], the results can represent the conditional binding constant between DOM and Cu2+. It uses Parallel Factor Analysis (PARAFAC) to model fluorescence excitation emission matrices (EEMs) of humic acid (HA) samples titrated with Cu2+ to resolve fluorescence response of fluorescent components to Cu+ titration. Meanwhile, Cu-ISE is employed to monitor free Cu2+ concentration ([Cu]) at each titration step. The fluorescence response of each component is fit individually to a nonlinear function of [Cu] to find Cu2+ conditional stability constant for that component.Conventional fluorescence quenching models including the most up-to-date multiresponse model have a problematic assumption on Cu2+ speciation, i.e. an assumption that total Cu2+ present in samples is a sum of [Cu] and those bound by fluorescent components without taking into consideration the contribution of nonfluorescent organic ligands and inorganic ligands to speciation of Cu2+. This study employed the new approach to investigate Cu2+ binding by Pahokee peat humic acid (PPHA) at pH values of 6.0,7.0 and 8.0 buffered by phosphate or without buffer. Two fluorescent components (C1 and C2) were identified by PARAFAC. For the new quenching model, the conditional stability constants (logK1 and logK2) of the two components all increased with increasing pH. In buffered solutions, the new quenching model reported logK1=7.11,7.89,8.04 for C1 and logA2=7.04, 7.64,8.11 for C2 at pH=6.0,7.0 and 8.0 respectively, nearly two log units higher than the results of the multiresponse model. Without buffer, logK1 and logK2 decreased but were still high at pH=8.0 (logK1= 7.54, logK1=7.95), and all the values were at least 0.5 log unit higher than those (4.83～5.55) of the multiresponse model. These observations indicated that the new quenching model is intrinsically sensitive than the multiresponse model in revealing strong fluorescent binding sites of PPHA in different experimental conditions. The new model was validated by testing it to a mixture of two fluorescing Cu2+chelating organic compounds, i.e. salicylic acid (Sal) and L-tryptophan (Trp) mixed with one nonfluorescent binding compound oxalic acid titrated with Cu2+ at pH=5.0; the fitting results of Sal and Tip (logKSai =3.02, logKTrp=3.71) were close to the theoretical values of 3.15 and 3.72 reported respectively in the literature. Besides, we also conducted copper titration to natural water samples collected from Daliao River and its estuary and tested our new quenching approach to the natural water samples. The PARAFAC approach showed that in the river-estuary system there existed 6 fluorescent dissolved organic matter (FDOM) components:Cl (ex/em= 279/310 (425) nm) was tyrosine-like fluorescence, C2 (ex/em=303/340 nm) was tryptophane-like, C3 (ex/em=310/380 (530) nm), C4 (ex/em=325/424 nm), C5 (ex/em= 380/485 nm) and C6 was humic-like substances. It was found that fitting the new quenching model to the FDOM in the seawater resulted in worse results than the fresh water samples, which may be attributed to that fluorophoric groups are in more complicated medium in seawater than in freshwater samples and thus the interaction between Cu2+ and those groups may not be accounted for by the same processes as in the freshwater samples.