Assessing the multisite binding properties of multiple sources of dissolved organic matter at nanomolar copper concentrations using piecewise linear regression and parallel factor analysis of fluorescence quenching

This study reports on the development and application of a piecewise linear model for the determination of copper-binding parameters at concentrations in the nanomolar range using fluorescence quenching. l-Tyrosine, Suwannee River natural organic matter, and two leaf leachates with similar fluorescence signatures were used as test compounds, and results were compared with those of the standard Ryan-Weber model. The piecewise model was also applied to and compared with data from an earlier study. Parallel factor analysis (PARAFAC) was used to identify three to five independent fluorophores in each test compound, and copper-binding parameters were estimated for one to three binding sites for each fluorophore. The binding properties of similar and different fluorophores were also compared. The conditional binding strengths (log K') estimated using the piecewise approach were similar to those obtained using the Ryan-Weber approach (p > 0.05); however, the piecewise linear model provided superior results compared to models based on the Ryan-Weber equation in several ways, including (1) capable of distinguishing more binding sites for a single fluorophore, (2) capable of extracting binding parameters at environmentally relevant, nanomolar concentrations of copper, where fluorescence changes are often observed as enhancement, (3) greater precision over repeated titrations, and (4) no severe underestimation of complexing capacities. Finally, the copper-binding properties of PARAFAC components with similar optical signatures were found to be similar, both in sources with dramatically different and similar total fluorescence signatures.